Oh this was brilliant! Thank you so much for your comprehensible explanation of technology related issues, especially concerning Horizon and players in the scandal! You know, if and when there are any prosecutions arising from the scandal I think your videos should be played to the Juries in Court! It would help them and others to understand what went so badly wrong! 🙋👍🙏⚖️🙏
I worked on a very large government IT system that was 25 years old. I saw long serving IT people, who had been there so long that they were "untouchable". There are two power structures in this type of IT organisation. The official management structure, and then the IT staff who been around for decades. They have so much undocumented technical knowledge in their heads, that the management dare not mess with them. I think that Gareth Jenkins had a lot more power in the company, than it looks like on paper. I suspect that if management wanted anything done, it had to be done the Gareth Jenkins way, or it wouldn't get done at all.
Ironically, it was these sort of people who were pushed out the door when the inhouse PFI Contract bid was put together. As they had the highest salaries due to their very long length of service. So they were offered very generous early retirement packages. To reduce the total annual salary bill in the PFI contract bid.
@@ChrisM541 The problem is that they have to maintain their status as the supreme company IT font of all knowledge. So will never admit to ever being wrong. Or to not have an instant answer to any technical question.
I've encountered these rusted-on IT experts, too. I managed to get rid of the most obnoxious one a few weeks after taking over the team. That gave the others the clear message that no one was indispensable and a culture change was mandatory. Eventually the old system was replaced and the remaining 'gurus' were made redundant and pensioned off.
@@James_Bowie That's NOT quite what's being talked about here. You don't go blanket-culling(!!) experience like that! That's utterly shocking and totally damning on you and your management team.
When we finally heard from Gareth Jenkins, I thought he was going to be an awful person… but he wasn’t. He had his witness statements altered by others, he had a pretty one track mind - great in IT but knew nothing about law - possibly on the spectrum - but wanting to get things “right”- I put that in quotes because he was told his witness statements weren’t allowed to include mention of bugs errors and defects - so he thought ok - they want me to leave that out.., The second sight guys had a breakthrough with their investigation because Gareth Jenkins told them all about the bugs errors and defects - because they asked him - so in a way, this whole house of cards came down because of his honesty - Ian warmington from second sight said he was one of the most honest people they talked to
I am so sorry for taking so long to acknowledge your comment. So good that you have reminded us of this Doris. Yep, Second Sight thought he was one of the good guys and their judgement was pretty spot on for most of this. p
Thankyou Paul for disecting the behaviour of these witnesses . Jason Beer is disecting them with his sharp mind huge research obviously to come up with such focused questioning
Thanks Paul - NZ here. Things I noticed: When NZ flight NZ901 crashed on Mt Erebus the airline was owned by the government and a cover up ensued. The public enquiry described the cover up as an "orchestrated litany of lies" to mislead investigators. The videos of the witnesses have similar "I don't remember" patterns as the PO Inquiry. Government run/owned businesses seem to have a weaknesses in this area - your thoughts on why this might be would be interesting. My background is software, network engineering, business management and also formal psychology study. The contract between POL and ICL Pathway and later Fujitsu was for POL to provide the spec and Fujitsu coded - it's common in industry. Fujitsu didn't have suitable software so they developed one based on what POL asked for - likely Fujitsu had limited experience in this area. POL could not provide a suitable spec as they were not competent to define a software system; postal mail companies are still very low tech, this was also part of where people failed. POL may have tried to replicate the paper based system they were familiar with - but paper systems have some inherent security and audit features. POL used a contact that outsourced all risk to the SPMs, so there was no incentives to solve Horizon issues - they just charged everything back to the SPMs - so no losses and no questions. Had POL brought in one of the big four accounting firms to work with Fujitsu, a better and auditable system that protected users (SPMS) could have been created. I suspect this is how govt control to keep spending down factored in. The other angle is that all software has bugs, this includes all the banks and retail chains, but they manage the liability differently and so solve the issues rather than pursuing the staff to recover missing money errors.
Well said, not investing in the computer system at the start was where this all went wrong, from a technical level. To replicate the paper system was a big mistake, a completely new IT system, based on POL’s requirements should and could have been designed. However vast amounts would have been required in training the post masters to operate the new system.Similar to Boeing Max scenario, hide the faults and live with the crashes.
I recently watched a you tube presentation on the Erebus crash. Essentially someone somewhere sometime altered the on board co-ordinates for a possibly genuine reason but failed to tell the crew. This was covered up and everything blamed on pilot error..Very apposite
Fantastic post. Thank you. Yes, government bodies do see to have a weakness here. I wonder if it is about authority structures combined with the value of duty - executive wield their power in the delusion they are serving the public and it becomes a kind of vocational zeal that results in real harm when they go off the rails. You can't impune their judgemetn because are a public servant and already morally impeachable. They think of themselves as self-sacrificing even though they are been very well recompensed for their work. Just some loose thoughts in response to your excellent question. p
Dr. DUCKET Your calm, logical argument I find inspiring and something l am trying to instil into my own approach to discussions . You are one of a small number of contributers, I follow, who have the ability to make the listener think and arrive at the solutions themselves and not to take things at face value. I thank you for that! It may stop me from being drawn into heated arguments.
I've been looking forward to your analysis of Gareth Jenkins. Your assertion of the morality of the coders being present in the code is fascinating. I would extend it to include culture. I've worked with Russian, Indian, British and American coders, in particular. When I reflect on this, I have seen overwhelming evidence of cultural and individual by-products in their code, and these are nearly always highlighted when we've had to fix bugs. I'm also interested in understating more on your thoughts of Gareth Jenkins himself. As someone diagnosed with Asperger's in later life, my subsequent research, reflection and understanding of the condition made me challenge my perception of Mr. Jenkins during his testimony. Whilst not to absolve him of any perceived ignorance or particular decisions, I hypothesised whether he had a flavour of autism which could have caused him to interpret, behave and act in certain ways throughout his work, and indeed, life. In some ways I abhor the use of the phrase "Common Sense" as I don't believe the term would exist if it were indeed "common". If we take for granted that Mr. Jenkins should have realised and thought differently then are we ignoring neuro-diversity? Society can't cope with trying to adapt to endless diversity, but could we better execute our work lives by structuring teams and assigning work in order to catch *some* of the more potentially impactful effects of neuro-diversity? In this case, I would propose that when legal teams and/or companies are preparing witnesses or employing technical architects and coders, that they apply more checks and balances on what will be the nature of the output of the people they are employing and how could they ensure a greater coverage of their needs rather than a narrow and individualised one. Unless, of course, they are choosing to use certain people in certain ways... How do I feel about Gareth Jenkins after all this? I perceive that his contribution to where we are with this scandal was less deliberate and intentional than the contributions of many of the other witnesses we have seen. In some ways I almost felt an element of sympathy for him in that he couldn't comprehend how many of his actions had contributed to the scale of the issues. That's not to say that many of his answers or actions weren't prepared, cold, ignorant or selfish. On the scale of guilt, I see others as more culpable, and maybe even guilty of manipulating him and his character. However, he is responsible for understanding his responsibilities as a witness. Whether one is or is not Autistic, one cannot use it as an excuse in all circumstances.
One of the other things which, in my view, clouded his evidence was the possibility that he was, to a certain extent, possibly coached by his own legal team. I felt, at times, that he adopted a different tone. I wondered whether he did use some ideas which had been suggested to him by solicitors or barristers when speaking about some aspects of his work.
@@mikejames4540 Agreed. Some of his answers certainly suggested this. How any of the witnesses that repeated their "I cannot recall" answers can believe they would be seen as authentic is beyond me.
Great post Chris, you make some excellent points. As to whether Jenkins might be neurdiverse, in one sense we all are, just some organisations take account of that and removed barriers to our engagement and perhaps other organisations exploit that diversity in ways that aren't in the interest of the person or others around them. p
Fujitsu turning a blind eye (with a view to minimising product liability) may be the crucial observation, because it leads to the question how Mr Jenkins was selected for his role of expert witness of fact (but, if possible, not an expert witness). It is hard to avoid the thought that he combined deep knowledge of the Horizon software with almost complete personal blindness as to the meaning and consequences of the Horizon faults and of his own participation in prosecutions. He was the ideal intelligent idiot, so it would not be surprising to find he had been chosen just because of that. Moreover, as his evidence to the Inquiry unfolded, he turned out to be someone whose evidence in the prosecutions was easily deniable, in the sense that he never had first hand knowledge of the corporate processes surrounding the Fujitsu contract. And, if he could be offered as a witness of fact, he wouldn't have all those inconvenient extra obligations to the Court. So our sympathy for him is engaged, as it is for any other patsy. But his oblivious attitude then demolishes our sympathy (if not our credulity). As for the morality of Fujitsu's approach to coding (and of the Post Office's attitude to procuring the system) this is, if anything, worse and more cynical than it has been portrayed. The system was not robust; it was fragile. It was expected to produce flawed output, probably for some time after it was rolled out, and consequently equipped with various ways of fudging the output from a safe remote location. It may have been hoped that bugs would be few and easy to spot, but there was a safety net for Fujitsu, that would operate at the expense of the postmasters. The moral issue for Fujitsu is that it was quite happy with this as a basis for delivering a turnkey system. It is not clear that Fujitsu expected the Post Office to prosecute the postmasters when Fujitsu fudged the branch accounts; probably Fujitsu was initially as surprised as anybody that this happened, once a week for years and years. When that happened, it turned to the oblivious Mr Jenkins. For Mr Jenkins this was an unending nightmare, clearly. The only way he could cope with it was by denying the fragility of Horizon and thrusting out of his consciousness the impact of his brief career as an expert witness. The limitations of his memory were surely as much real as they were deliberate; he has had to spend 15 years preserving his own sanity. That does not excuse, in any degree, the way that he failed to take elementary steps to verify the sweeping assurances he gave about the robustness of the system, rarity of hardware failures and so on, when appearing before the Court. But that oblivious lack of care was what one might expect from such a carefully chosen stooge. The greater guilt lies elsewhere.
Yes, if you look back at the evidence from Jason Coyne (v early expert tech witness, day 59, was expert in 2004)he highlighted issues, possible errors and discrepancies could be caused by Horizon. POL tried to get him to change his evidence, and when he wouldn't, got rid of him. You can imagine from that how pleased they were to find GJ, and how keen they were to continue to use him.
@@gailb8722 Absolutely! But I think it was Fujitsu that pulled the relevant strings to find Mr Jenkins, later on. The Post Office here was merely incompetent. Even Jarnail Singh seems not to have been making a clear-sighted choice, taking what Fujitsu gave him, when he was in a tight corner.
This is such a good post. Really powerfully argued. Thank you so much. It is a really valuable contribution to the discussion and I think there are som many good arguments in your careful scripted post. Thank you so much :) p
Exactly correct: having suffered at their useless hands for the last two years I can but agree, It is not more money that the NHS needs, but some solid management. Overstaffed, underemployed, overpaid. GP's are even worse: eight weeks to get an appointment, even though just discharged from hospital and needing further medicine - they are but a lazy joke.
Wow! Super glad, I found your channel. I have been following the Public inquiry (also from Australia) partly because I lived in the UK from 1999 - 2007. I find the inquiry as a whole, fascinating and engrossing. I enjoyed your video and it has opened my eyes to yet more insights that I not even considered. Thank you.
I strongly believe this whole issue was a people and culture problem. Horizon is no different to any computer system - Jenkins is right about that - they all incur bugs and defects. It was the persistent culture of denial - primarily to protect the brand that is at issue.
True, but the Horizon system did seem a really bad system, many of the postmasters demanded to see their transaction records in order to prove their innocence, amazingly these records were/are only available for a month before deletion, contrast that with say Amazon, where every purchase you've ever made can be called up instantly, or your banking records, ditto, the whole post office culture was to preserve the facade of software integrity at all costs to please the government and shareholders.
@@monteceitomoocher 'many of the postmasters demanded to see their transaction records in order to prove their innocence'. This is the crux of the matter: we have heard again and again that the postmasters 'offered no evidence' that the Horizon system was faulty. How could they, they had no access! Besides in British law you are supposed to be innocent until proven guilty - somehow the burden of proof got reversed in these cases and that is a defect of the legal system. Astonishingly cases were allowed to proceed all the way to conviction without the Post Office even having to prove that any money was actually missing let alone that the accused had stolen it!
@@john_hind Yes. But the position of the P.O. counsel is far more culpable than you have indicated. I am informed that the Fujitsu equipment -- (either at head quarters or elsewhere) -- had a (permanent?) record of EVERY KEYSTROKE entered by a SPM. This record would provide CONCLUSIVE evidence of any misappropriation of funds. Why was this evidence not made available -- in the case of a criminal trial? Who made the decision to withhold that evidence?
I was at the Inquiry for his evidence from Wednesday to Friday. It was fascinating. I agree he was a hard one to work out. I had a lovely chat with a lady who was also attending about what we thought of various witnesses and your videos came up as we'd both seen and enjoyed them. Just thought you might get a kick out of knowing your videos were being discussed at the Inquiry 😂
Well, you're right. I got a tremendous kick out of that. Thank you so much for letting me know. That has really given me a boost. And, well done for going to the Inquiry. Having people like you there keeps them all accountable :) p
So I used a computer system designed in 1989 in a major uk utility firm. Everything i did had a used ID and time stamp. Even if I looked up a customer or an address it was recorded.. After a shooting , a number of curious people looked up the shooter, and got reprimanded. The idea that you could change a balance without a data stamp and user ID is madness… Even if the system changed a balance. The system should assign its on ID.
Totally concur and that balance transaction date/time/User ID stamp would have another date/time/User ID stamp against it when it was "corrected/adjusted" by a Fujitsu "adjuster" showing it had been, in these situations, been "tampered" with.
Yes, but you were an END USER.... just like the spms were. *Fujitsu own and have access to the CODE fgs!* Of course they can do this. It's a no-brainer. Single entry adjustments were made by Fujitsu engineers directly into the backend. What was missing was the appropriate security/governance/signoff processes and procedures. When Jenkins sent his witness statement to Jarnail Singh, he also sent a memo in attachment. In it, he described three possible solutions for branches affected by the receipts and payments mismatch bug. TWO of the three options involve doing PRECISELY this. *As he described it on the second day of his appearance at the Inquiry: 'it would involve changing Branch data without informing the branch'* The memo also speaks to the RISK of each option and the moral implications 'to the Integrity of the business' of manipulating data in this manner. THATS part of the problem with Mr J. Not only did he suggest and describe this in that memo - which we clearly saw in Mr Singh's testimony AND Mr J's - the memo was an attachment to the same mail he sent Singh his witness statement, in which he said: "No external systems can manipulate branch accounts without user awareness and authorisation." -- Gareth Jenkins witness statement 2010, for the Seema Misra trial. *He knew this was not true.*
Thought provoking video - I have always maintained that any business should be an honest endeavour based on fair dealing and benefit to all concerned. Larger corporations seem to become arrogant and Machiavellan over the years treating many of their staff. customers, suppliers etc as useful but disposable but the PO scandal seems to have shone a light on this assertion. But the question is why do so many at PO seem blind to their lack of deeper insight or their own morality ?
Appreciate the balanced view , perhaps if more people like your goodself had of been involved we just might not have been in the position we are in , keep up the good work
I found it interesting that Jenkins was still very defensive about Horizon. But most notably to me, he couldn't adequately explain why he had repeatedly failed to reveal important information that would have pointed to defects in the system.
It was his baby after all - but I think he had been admonished after a very early court case by POL or Fujitsu to not reveal Horizon inadequacies so openly in court again, so being a company man, he did not..
Dear Dr Paul… The thing that struck me about Gareth Jenkins’ testimony was about the standard Fujitsu witness statement by himself and Penny Thomas and Andy Dunks that the computer was “working correctly”. Gareth pointed out this wasn’t the Horizon system/computer but was the computer used by them to extract the audit data. I thought that was a significant point that I had not appreciated or considered previously, and I don’t think the Inquiry had considered either.
@@cassandratq9301 Thanks for the replies. Gareth said that Penny Thomas and Andy Dunks would have no knowledge as to whether Horizon was working correctly, but that the 'computer working ok' in their prosecution witness statements was a reference only to the computer used to extract the ARQ Data from the audit data base. Andy Dunks is giving further testimony today - so it will be interesting to see what Andy has to say on this.
This line of questioning to a number of witnesses both confused and irritated me. As I see it the original clause (drafted by lawyers in both Fujitsu and Pol) referred to the service by which Fujitsu extracted transaction data from the archive and presented it to Pol, and was meant to certify the validity of the data and the function of the extraction mechanism… “the computer”. later Pol was in the habit of drafting witness statements for Fujitsu and used the same template, the clause had no place in a witness statement which did not include transaction data extraction but was still included. Hence Jenkins’ assertion that it must refer to the pc that he typed the statement on. The inclusion probably suited Pol because it was unclear. Later still, by persons unknown, “the computer” became “the system” which gets Mr Dunks and others into a world of trouble and suits Pol even better. I don’t believe that the inquiry does not understand the provenance and timeline of the witness statement pro forma and it irritates me that they use its confusion to attack witnesses when they should be searching for the truth. My take on Jenkins…. A very busy tech “beard” trying to be helpful and trusting the legal system to look for the truth, I take you to the testimony of the prosecuting barrister In the Castleton case - “it’s an adversarial system and I won”
thank you very informative. To me there are two angles to this scandal which are related. Firstly the technical lack of moral code which is highlighted in the video. Secondly the commercial, legal and career preservation lack of moral code(s) displayed by the PO management at all levels.
“Oh, just one more thing,”said Lt Columbo. GJ worked his entire life at ICL (ignore the name change to Fujitsu for now). By the time it got the post office/benefits agency gig in 1997/8, it was a busted flush. Its only customer of note was UKHMG. It was British (huzzah!) and it was cheap. Its technology was poor. In many cases ways it resembled British car production, traded on former glories. Just thought that’s worth remembering in all this. I can’t wait for the ‘Horizon’ movie!
hahaha, nice one :) I think tech companies have a habit of dumping bad code on governments and the like. I see it all the time in education - until Google and Microsoft woke up to how schools and colleges had children who would grow up to be tech users and so needed to get them young and get them on brand, techn companies used to just give schools crapware because they weren't seen as 'enterprise/business' and usually didn't complain.
@@PaulDuckett Thanks Paul, good insight. British exceptionalism and our fetish for flag carriers doesn’t help. I worked for an offshoot of Cable&Wireless, aka the telecoms branch of the Foreign Office. Boy, did the empire live on there. The C&W chairman had a butler. The Mercury building in Bracknell was opposite ICL hq (early 1990s), car park full of range rovers, range rovers full of golf clubs. They were already crap, then Wang and Dell arrived, almost like mourners!
Thanks, very interesting. For me, the "high level" analysis would also include alienation - the alienation that results from an economic system that prioritises financial profit over any other socia/human outcome of activity, and the alienation of individual workers located within a corporation so large that they have no contact with the people their work affects
Thanks for your thoughts. The moral ingredient of computer coding is surely one of the least discussed and yet possibly one of the most important. Your questions are very salient and quite disturbing. Thanks again for your calm and well-articulated analysis.
How interesting about computers and moral code. We spent a couple of days at the Public Inquiry and came away with a very similar feeling to yours about Gareth Jenkins.
Sorry for my late response. Great to know you are attendin the hearings - it's so important that the public fill the public gallery. Great to know we saw this in the same way. p
Dr Paul, thank you for your analysis. Tying the Horizon scandal to the "Move fast and break things" mantra of Facebook and the wider IT industry is an interesting thought, particularly when recognising that the mantra had its genesis in software development practices such as Extreme Programing and Scrum methodology, which both date from 1993. To give him his due, Mr Jenkins was not the design lead architect, but seems to have inherited the title of "Mr Horizon" some-time around 2003. Too many of my colleagues fail to make the intellectual leap, that software is not morally neutral, and that we have a duty to Society to do all that is humanly possible to produce software (and attendant manual systems) that act as good moral agents.
I am so sorry for taking so long to acknowledge your comment. Thanks so much for sharing your thoughts and I think you are right to give Mr Jenkins his due that he didn't design that thing, he just got sucked into it all. p
As a former contract manager I have always hated the adversarial form of contracts, rather than a collaborative style. I know it’s very hard to achieve. It does not serve either party when severe and threatening clauses make it impossible to be honest with the buyer about issues. This clearly happened here, the PO screwed the price down so much that the operational side couldn’t afford the testing and monitoring involved. They couldn’t speak openly about issues and it encouraged them to lie about how the system was working.
@allisonstandley2573 Jenkins would say he didn't lie, or at least didn't mean to. He appears to have been carried along into agreeing with statements that PO wanted, and he was unwilling or unable to prevent them Whereas PO seem to be institutionally corrupt, Fujitsu seem to be inept and unable to define a robust position for themselves
Yes, I agree. It's those corporate contracts that can be so horid. PO was supposed to have a relational contract (one premised on negotiation and discussion), but they didn't do that which is what caused such horrid outcomes. p
I agree. A significant reason that I left Fujitsu in 2001 was that I hated this kind of project, which accounted for an increasing part of FJ business. You end up with customer and supplier constantly trying to score points off each other rather than working together to a common purpose, which I have seen happen to great effect in private sector projects of similar scale. It's inevitable with this kind of large project that the last few bugs will be found during live running, and that isn't a problem if everyone recognizes that and works together to deal with the fact. Openness with the user community - sharing information with them and encouraging them to share information with each other - is a vital part of that, and is what was totally missing on this project.
It was interesting when Mr Jenkins said that he was better with computers than people. I have the feeling that would apply to many people drawn to working in IT. At times he appeared utterly bewildered and unable to understand why people were so angry and frustrated by this situation.
Thanks Dr Paul, very interesting. I was struck by the way that in his reports to the court, GJ answered very narrowly the questions that he was asked. In his oral evidence he said he hadn't considered the wider picture, that no one had asked him, and I believe him. If you look back at the evidence from Jason Coyne (v early expert tech witness), he highlighted issues, possible errors and discrepancies could be caused by Horizon. POL tried to get him to change his evidence, and when he wouldn't, got rid of him. You can imagine from that how pleased they were to find GJ, and how keen they were to continue to use him. The way GJ struggled with open questions, and answered tech questions with rapid very detailed enthusiastic answers seemed like something I've seen in many software engineers and similar. As GJ said, "Better with systems than with people". Thanks for your interesting analyses!
@@cassandratq9301 Interesting question! I would say yes, and Post Offices's behaviour of keeping using him and not risking finding another independent expert!
Sorry for my late response. really good reflections. I didn't spot the narrow answers, but I think you are right. Thanks for sharing your analysis. Really worth reading. p
Horizon was shit but wasn't the problem for the SPMs. The problem was the ignorance and fanatical nature of POL corporate governance and the pervasiveness of some of hateful social attitudes.
If horizon had been perfect then the issue would not have happened at all, but of course no software is perfect. If it had been good enough then they would have had occasional issues that they could investigate. This didn't happen either and the prejudice was allowed to take over You can't prevent prejudice, so I would argue that horizon was the main issue.
First, I really love your channel, your content is always thoughtful, thought-provoking and informative. Second, your take on Gareth Jenkins chimed with my own. He seemed to see things in quite simplistic terms and I even wondered at times whether he could be on the low end of the spectrum. He looked through documents by ignoring anything that he thought didn't apply to him. He did what was asked of him. It's evident he was vulnerable to being used by people who had an agenda to push. This doesn't, however, remove his culpability and his apparent inability to appreciate the human cost. Maybe you could say he knew how technology worked, not humans. Thirdly, to me this video didn't seem too long at all. In fact when you said a couple of times that it was coming towards the end I felt quite dismayed, and said out loud "no! I could listen to this all day". Thank you. Look forward to the next.
Thanks so much for your lovely feedback and your thoughtful post. Yes, the question of Jenkins' neurodiversity is an interesting one. I think we are all neurodiverse in one way or another but that some forms of diversity are exploited by companies like Fujitsu and POL to cause harm. I think that Jenkins' might have been exploited in that way. Perhaps that is why we couldn't blame him as much as we thought we could. p
- your comments made me think about the ethics approval I needed for my qualitative research and you're right, once ethics approval was granted I implicitly understood my research to be ethical. Perhaps even moral to some degree. Thus i paid it no further heed. I don't know of the any ethics approval or overseeing body which regulated the Horizon system. Shirley future court cases will focus on this failing. All of the wretched individuals in this sorry episode and injustice, to my mind, have essentially minimised their individual involvement and culpability. 😕🥺😕
Excellent point James. Yes, ethics has to be continual rather than episodic. I think a lot of folk get caught out by that. Sorry for taking so long to respond, your post was excellent, I'm just very tardy!
Fascinating and thoughtful in equal measure.I spoke to GJ's KC prior to his evidence and also Fujitsu(they are not representing him) and suggested that he was in my opinion incurious about the detail of matters outside his work/interest,typical of Senior IT people..She replied "lets wait and see2.The question for Fujitsu is"How many other Dept. did you have that supplied prosecution support to bring end users to court?
I just want to know it was determined which postmasters accounts were corrupted and how the amounts were determined.... If it is true, that the postmaster who actually saw the staff was then targeted to the tune of 40 some thousand pounds, there is a record of that in correspondence email or paper or bank accounts... The code allows the data manipulation/insertion. but somebody is telling them which accounts and how much....
This 3 day interrogation of JG had many commenting on Live Chat perplexed too. It has been suggested several times and on different days that GJ is on the autism spectrum. The reasoning being - it was suggested - that this would account for the detached way in which he spoke of events (note: not in an ‘aloof’ way but ‘detached’ and ‘coldly objective’). It was also noticeable that he liked the technical questions from the change in his general body language and demeanour to his interrogator. Concrete questions good, value judgement questions bad. This meant (it was further suggested) that he could be taken advantage of and by the lawyers in particular as he judged these important people by his own standards.
So sorry for my very late reply. I am still catching up on the comments. You make some really good point. I didn't realise that was happening in the Live Chat. Gosh, very interesting!
How many of those responsible will face a criminal trial and prison? The Post Office inquiry is theatre to appease the masses. We saw with Partygate that laws are for the little people.
And very expensive theatre it is. Done better to get back all the bonuses, not spend millions into barristers pockets with this droning on enquiry and pay the Sub Postmasters substantial amounts that is tax free and their legal fees paid by the post office.
As someone who has worked in the nuclear safety business for more than 25 years, I have always believed that an ethical dimension is an essential element in our work and MUST be explicitly identified. Unfortunately, like in the post office, this has been treated as “irrelevant” to what is treated as a simple technical issue. Thank you for your thoughtful and very accessible broadcasts.
Ethics are not included in your employment contract, in fact they are abrogated by any NDA you might sign.... Imagine that... You think you will get a bonus for your ethics but that is purely delusional in the current culture...
Paul, I think that while your argument about the morality of code is thought provoking and a valid one it is a stretch to apply it to the Post Office Scandal/Horizon. It’s a system working with financial transactions - the scope for morality in the coding is very limited (if it’s there at all). I can only see that it creeps in in a very human and insidious way in terms of the coding allowing access without notice to the sub postmasters. The overriding point has to be the human (inhumane) reaction or inaction to the flaws that were there.
I have to disagree. My version of properly moral code would have had the back-door inserted transactions clearly identify who made the (ahem) "correcting" transaction, at what time & day, and from where. The Horizon system had none-of this. ergo: a purely technical financial, yet morally compromised solution.
Thanks Goffers, you make a good point. I may have over extended my argument. But glad you engaged with it nonetheless and posted your thoughts. I really appreciate that. p
You dont understand tech, that isnt what is happening and cant be done in financial systems in the UK . Yes all software has faults but this company is unregulated unlike every other financial business. They prosecuted people for stealing and didnt return the money to where it was suposedly stolen from. Gareth lied in court repeatedly and then after the PO were told not to use him anymore they still did and he co-operated throughout. They found someone who was prepared to keep saying what they wanted in court despite evidence to the contrary. The software broke ACID rules, he gave evidence on the quality of the code but couldnt answer Mr Beers question if there were unit tests in it. Not that would have saved the errors but it proves he hasnt looked at the code.
Thanks for posting. You are right, tech is not my area of expertise. It's good to hear your different take on this. Thanks so much for sharing as you've made some good points, p
I think there are several factors that might explain Gareth Jenkins contribution to the Horizon scandal. First, he appears to be what we now call a nerd or a geek. He is obviously an accomplished IT engineer - hence the Distinguished Engineer title. People who are very good with numbers (mathematicians, statisticians, accountants, data scientists) are sometimes lacking in another critical area - words. These people often feel more comfortable processing numbers (calculations, equations, coding, etc) than talking to others or expressing ideas or opinions unrelated to their specialisation. To his credit, Jenkins answered all questions put to him in the inquiry patiently, clearly and (surprisingly) concisely. He considered every question carefully and gave accurate and consistent answers. It is clear why Fujitsu chose him to be the expert witness. The second factor that may help in understanding his behaviour is loyalty. Jenkins had essentially worked for the same organisation his whole professional life. To use a hackneyed phrase, he was a company man. He owed them. The third factor is pride. He had the Distinguished Engineer title bestowed upon him by Fujitsu. To criticise Fujitsu or Horizon is to criticise himself. Also, he was instrumental in designing the Horizon system - which he continued to defend throughout the inquiry. To accept that the system was faulty meant admitting that he had failed in his role or had been negligent. Neither was palatable. From his standpoint, the system worked as it should. BEDs were at an acceptably low level. The only problem was that the consequences of those BEDs were convictions of real people. And "people" was an unfamiliar concept to Jenkins.
@@cassandratq9301 Having said what I said, I believe he does bear responsibility for the evidence he provided in the trials of SPMs. He must have understood that the evidence he gave would likely result in successful prosecutions.
It seems that the public and various bureaucracies have outsourced their morals and ethics and so they say, ' Not my decision, idea or whatever'; not my fault'!
I'm loving your take on things. The morality of coding seems so obvious they way you present it, but yet it was absent in so many people's thinking. Maybe because it can't be measured and there's no expectation for accountability? Me? I'm just sick n tired of inadequate corporate processes. It seems everything is designed to go wrong, and does so often.
Hi again from NZ Paul, thanks for your video and if you hadn't mentioned yr own laptop issue, I wouldn't have noticed - all the same you are very clever with your editing tools. You raise very good points here about the process of software (SW) design, development, testing and eventual release to end users and its potential impact of unintended consequences (the social/moral aspect). My SW background is mostly in military and civil aviation and the consequences of getting it wrong can be catastrophic of which the Boeing 737MAX is sadly an example of. In the case of Gareth Jenkins, and no disrespect to SW engineers/coders is intended here, he reminds me of a very small number of these folks I've worked with who think only at the bit and byte level, ie down in the weeds. When given a well defined spec with clear expected outcomes, most will meet the desired goals. However if the spec is poorly defined then the SW will most likely not achieve its desired outcomes and that is clearly where Horizon sits I suspect. This said, Gareth, like the others in the management chain here need to be made an example of as with what we've seen to date from the enquiry, the comparison I'd make is they've all been driving at speed in a car that has been running on only the wheel rims and they've chosen to ignore it simply by turning up the car stereo to drown out the noise in the hope it will be alright and go away. Thanks again for your videos, I find them entertainingly enlightening.
Sorry for taking so long to respond Brent. Just catching up with things after being away from youtube. You make some really important points and it's great to hear your views on this given your tech experience. Thanks so much for sharing. And, thank you for your lovely feedback too :) p
Watched hours of testimony from many of the witnesses, often make a decision on what I think they did. Then Mr Beer moves in a direction that goes against my thoughts. What I learned is that without all the information, we are likely to make the wrong decision on guilt. Only view I would trust is the final view at the end of the inquiry. No stone left unturned
My husband has just watched this video. He simply commented that Dr Duckett reminds him of an old sociology lecturer of his in the 1980s, Dr (later Prof) John Hughes at Lancaster University in the UK.😊
Another fantastic analysis, I worked in tech for 38 years , and was beset with "B.E.D's" and the end user had to find these and we as field engineers reported them back, customers lost faith and took their business elsewhere but ultimately found the same issues, so you're absolutely correct it is endemic in the tech industry. Ultimately GJ wrote that code without careful consideration to the consequences of his failure to write it properly therefore he IS responsible for those consequences but failed to acknowledge that . But it's not a PlayStation game that you just press reset and start again. No ...its immediate custody and ruin of innocent people's lives . Thankyou for these brilliant videos Dr Duckett 😊
There is a crisis in Academia , with thesis / studies using faked results to reached the intended goal . There was an old idea of a " No Blame " culture, this allowed admission of mistakes .... we need to revisit that idea
Rather a simplification, I think. Firstly, I doubt that GJ wrote much of the code himself. In fact, I doubt that FJ wrote much of it: a lot of the problems were that they selected the third-party Riposte sofware as a key central component of the system, and it proved insufficiently scaleable for the transaction volumes being handled here. The software did eventually get to the point where it was reliable enough to meet the operational needs of the business, the problem was that that's not the same as being reliable enough to be used as the sole evidence to prove that users have committed a crime.
I felt the same conflict. He came across as quite lame and was clearly filling a role he neither chose to understand nor was qualified to fill. Seems willful in his ignorance as he was enjoying himself too much. He didn't illicit anger (as some have, deservedly so) but he is definitely culpable as a willing participant. Great thought provoking video, as always. Thanks.
So sorry for my very late reply Stewart. I am still catching up on the comments. Really appreciate your comment and yoru encouragment. Thanks so much. p
I think one important distinction in the Facial Recognition comparison is that Horizon was suffering bugs while the Recognition software was functioning as they'd designed it (but that design was flawed). when Horizon added things to balances when it shouldn't be, it was performing actions that shouldn't be part of the process. the FR software is, for want of a better term, "working as intended" but it's just bad at its job. the racism coded into it means it does poorly on black faces, but that just means it spits out the wrong score/result for its comparison check. it's a flaw, but it's not causing unintended behaviour; by which I mean it's not like when the software sees a black person then it suddenly starts trying to print the picture or ask for the user's login details. it performs poorly, but it still performs the actions it was told to perform. Horizon behaved in unexpected ways. it initiated processes and did things it shouldn't have been doing. so while the comparison is interesting (and I think has some merit in terms of thinking about the impact of code and being careful when you write it), I'm not sure if it's the most applicable
I think Jenkins genuinely thought of himself as a 'God of IT/Programming', and as such, took the stance that Gods take instruction from no one. I genuinely think he dismissed so much because he though it was beneath him/bothersome to engage. He even took on the title of 'Dr' with no objection...or correction. However, you do find these characters particularly in computer programming, a field where the socially awkward can safely gravitate. What he doesn't seem to grasp is the significance of software bugs/defects on the 'users'. I bet the film Tron sits high in his favourites list. He was most definitely used to the maximum by the PO's legal team, clearly 'allowing' him to leave out crucial evidence, or rather, informally instructing him not to 'disclose'.
Great post. Thankfully at my place of work our IT folk are fabulous. They don't they they are gods but the rest of us do. But I completely recognise what you say. It's when knowledge is used a authority. That can be really risk as knowledge is often flawed an incomplete. p
I, too, am conflicted in my opinion of GJ. Gareth Jenkins was one of 30 developers at Fujitsu, and 30 shared service centre staff. There were many layers of Fujitsu above him, and many layers of Post Office Employees above Fujitsu. Whilst it's right that GJ's part in this miscarriage of Justice is scrutinised, he is a small fish in a big pond, and it would be wrong if he was the only one so heavily scrutinized. He was remarkably blasé about his evidence provided to POL and how crucial it was in convicting many sub postmasters. This suited POL and Fujitsu, at the time, and now. His part in providing evidence seemed to evolve over time (boiling the frog analogy) , and he wrongly, never seemed to grasp that he was an expert witness with the same responsibilities as an external expert witness, and wasn’t just providing a couple of paragraphs for a corporate report to be used by a customer of Fujitsu. In his early evidence, he stated that Horizon was 5%- 10% of his time , so he had many competing objectives for his time. He provided circa 15 witness statements over a dozen years. His reports, commissioned for different reasons , were used many more times, both in court and by Senior POL management, but he was unaware of this until the Inquiry He was remarkably casual in providing blanket assurances, without making an effort to research if they were actually true at the time, and continued to be true, or just assumed to be true, based on informal office conversations, or that had been true when he first learned about them. To determine if these assertions were true , would have taken significant additional resource, and he could have refused to provide those assurances, without the extra time and resource. He didn't. POL solicitors, having received those assurances of certainty, from an expert witness, were happy to take them and use them to prosecute. It's clear now that the functionality, to insert transactions at Post office Branch level, was specified and built from in the start.(2003) It may have been intended (rightly) to be used rarely, with high levels of authorization, clear transparency and auditability. As it was actually deployed, none of these were true. Many people could use it, inserted transactions, were not authorized by Sub postmasters, and the identity of the users who posted the transaction could not be relied on to be absolute.The audit reports available did not show the crucial information, i.e. externally inserted transactions. The use of inserting transaction at central level, by many people , had become common place at the shared service centre, possibly because of faults in the software, and the need to fix them to allow business to continue at Post office branch level
Externally posted transactions were often identified as the last local user logged on (usually the sub postmaster), not the shared service centre agent or developer. The actual user could only be identified by extra work and analysis of base level transactions, none of which the sub postmaster had access to. The standard reports showed the "provided" user ID. Investigators rarely asked for the extra analysis, because they never knew if the User ID on the report was a problem. There was no incentive for them to ask for this, as it took time and money, and would only weaken any prosecution case. They received cash bonuses on successfull prosecutions. The sub post master was under intense pressure every day, to balance the books correctly or accept their culpability for the out of balance amount, or the branch could not be opened the next morning. It appears to be common place for sub postmaster s to be in the office until the early hours of the morning, and then having to put through a balancing line, or not open at 9:00 a.m. “Caught between a rock and hard place” for a small business owner. Imagine having to leave the shutters down and explain to the queue, "I can't open because I can't balance my books" When originally deployed , the out of balance amount could be posted to a suspense account. But the amounts on these accounts were often several thousands pounds. This was a worry, so instead of finding out what problem was causing the need for balancing amounts, they blocked the sub postmasters from posting a balancing item to the suspense account. This accelerated the values of unexplained out of balance amounts, and potential prosecutions The wider issue of the unequal contract between POL and the sub postmasters, encouraged POL solicitors to prosecute Sub postmasters to recover "missing funds", rather than pursue the root cause for the errors, which were more than likely problems with the Horizon Software, and therefore for Fujitsu to fix. The lack of scrutiny from POL to Fujitsu on the integrity of the Horizon system, suited Fujitsu commercially, as they were engaged in selling additional contracts to POL and other Government agencies throughout this period, so having a key platform branded as unreliable would have made winning new business almost impossible. So GJ as a useful idiot suited them Horizon worked well for most branches most of the time, but there were regular issues. These issues seemed to arise with rare combinations of events and were often difficult, time-consuming to replicate, and fix, and new ones continued to appear over time, with new combinations of hardware and software and operations. IMO this is pretty normal, especially for bespoke computer systems. Software company fixes them and life moves on. It doesn't usually end in someone going to prison
So sorry for my late reply. I think you post was really important . Thanks so much for taking the time to lay out the issues in this way. It's very valuable and I am very grateful. I think other's will benefit from the careful thought you put into this. p
@@PaulDuckett No problem GJ' was painted as a villain, by the media, so I was keen to understand how much he was to blame. After watching several days, it was clear to me, a newcomer, that it was more than a one person issue.
@@Pat14922 Indeed, it was more than a "one person issue". However - one critical aspect of GJ's testimony fails the sniff test to me. I find it incomprehensible that someone with the intellect to achieve a Maths degree from Oxbridge can claim not to understand or known (or research) the difference between a civil and a criminal case, especially if knowingly providing a statement (let alone an expert witness statement). On repeated occasions. Including appearing in-person in one of the seminal cases (the Misra case). To me, that's a line he was coached to take. There's clearly been massive coaching taking place for inquiry witnesses, to the point that much of it is useless. I speak as one who's done the IT expert witness bit just once in the late 80's, in a civil case. But then I'm a mere Imperial College engineering grad....
@@mpellatt Massive coaching? Agreed. It's written all over their testimony. The witnesses know a solicitor will be poring over their testimony on this inquiry in deciding who will face criminal charges. GJ - is quite likely to face a trial, and the court will then determine if he was ignorant or criminal My concern is that there are many more who could and should, but may escape the consequences,
As someone who has some IT knowledge, I used to build PCs for fun. I can tell you that trusting an IT system to work 100% correctly is ridiculous! Of course, it's going to have problems. A system of that scale built today wouldn't work 100%, let alone designed in the 1990s! As you say, its inability to record and report errors is a big issue. But so is the fact that the Post Office didn't seem to care that errors were not being corrected and, instead, blaming the Sub-Posts is the scandal. Jenkins' insistence that he didn't know what an 'Expert witness' actually was seems to be his only line of defence. So it's the 'classic' - Is he utterly incompetent (not knowing his court role) or a liar! (who knew his role and is now pretending that he didn't)
So sorry for replying so late. Excellent insight here. Yes, it's crazy we'd ever believe a computer would be 100% correct, given computers are designed by humans! p
Jenkins confidence in his own abilities is a classic programmer's spectrum response. His candid approach was as a result of being abandoned in the forest and assuming thinking he was walking on a path
Congrats on a thoughtful piece and promoting a thoughtful discussion. your “move fast and break things” alludes to ‘agile’ software development which you reasonably well describe some aspects of but which would not not have been used for Horizon. Nevertheless Horizon like 99% of IT projects, according the evidence, was accepted into service by the customer and according to contract with faults remaining to be fixed and further faults remaining to be found. Your assertion that the product was “released to market” is mistaken as is the suggestion that there was no mechanism for finding faults or fixing them, it is highly unlikely that any fault would have been fixed without testing and acceptance by POL and released into service under the control of POL. IMHO POL’s assertions that they knew nothing about bugs errors and defects Is simply absurd. I’m not sure that Fujitsu would have contracted for a system development whereby POL would prosecute its agents for failures in the system. I feel that Fujitsu were remiss in not controlling their clients actions (and that sounds stupid when you write it) but I cannot get past the actions of the legal profession in promoting and supporting this miscarriage, all the while professing to be simple agents of their client’s desires and simple mouthpieces of their client’s words.
I am so sorry for taking so long to acknowledge your comment. You make some great comments here John and you are right to correct me on those things. Thank you for pointing those areas out where my analysis didn't quite work. Great to hear your insights on this. p
My job was about 25% reducing the occurrence of harm to a a very small figure. When something went wrong, our first priority was identifying how to prevent it from happening again. But hey, what do I know, I'm a mere nurse and operating department practitioner.
I am so sorry for taking so long to acknowledge your comment. Great point. If only those folk worked to your standards. I think we seem to value those industries that churn out immoral actors and forget about valuing those sectors of society that only function because they are populated by moral actor. p
You make a lot of good points, but many of them only distantly related to specifics of Jenkins' testimony. I think you hit the nail on the head in your final sentences: FJ didn't design the system for the purpose of prosecuting postmasters, and when they discovered that their clients (POL) were using it for this purpose, they turned a blind eye. A lot of this is the culture of "the customer is always right", some of it underpinned by what's written in the contract: if the contract says you will support the client by appearing in court, then that's what you do. I don't think there was a "move fast and break things" culture here (you overlook that many of these events were 25 years ago, not 10 years ago), but there was a culture that involved lack of openness - sharing of knowledge with and between subpostmasters would have made everything operate so much more smoothly. The key causes of the problem in my view are (a) the adversarial supplier/customer relationship (which is normal in public sector procurement, and accounts for the fact that public sector IT projects are so expensive and unsatisfactory), and (b) the adversarial relationship between POL and the subpostmasters, which is a distinctive feature of this particular project.
Thanks so much for your posting Michael. It's really well thoughtful and I think you analysis is really useful. It very much adds to our understanding of things. p
I think that is a very kind, sympathetic and fair assessment. Mr Jenkins, I think, was strangely and, you might say choosily, incurious. He was poorly briefed by his legal people on the ‘expert witness’ role, yet seemingly failed ever to question his role, other than to say he thought he was on the prosecution side. He betrayed a certain vanity, for example, his pride over the ‘distinguished engineer’ badge. I think the reality is he was a lonely, long-time company man, and was easily flattered by the attention he got. He let it go to his head. But hey, enough of my yakking (Spinal Tap gag). Oh, last thing… please do a number on the government reps’ roles at UKGI and SHEX. Their culpability is striking, and will almost certainly go unpunished.
So sorry for replying so late Stephen. That is a great observation. Your word 'incurious' is perfect and might be applied to a number of the POL executive too. So, thanks for posting. If I can get to UKGI and SHEX I will. Is there a particular witness you want me to look at? p
@@PaulDuckett No apology needed. Thank you for replying, and more pertinently, for your excellent posts here on this and other topics. I've yammered on elsewhere about the PO's own 'one of us' mentality; I'll have to revisit the inquiry catalogue to remind myself who's who. From memory, it's easier to recall who not to finger, if you'll pardon the expression. Anyway, thank you again for your great work.
I am a super geek and have been writing and releasing software for 30 years, there is no way horizon would have been released if I was involved. Do you watch / read any Dan Ariely?
Thanks for posting and I hadn't heard of Dan. I had a quick look and he seems to be a really interesting fellow. I will try to read some of his work when I can. Thanks so much for the suggestion. p
He can and will say that he was never properly instructed and having seen Jarnail Singh provide evidence he may well be believed.I also think POL changed his statements without checking with him perhaps because it was obvious he was really engaged with the Legal Process.POL wouldnt put him in the box ever again but did use his evidence which was rather naughty.
As you say his testimony was hard to assess Overall, I would say he should have been nowhere near a courtroom, and did not have the skills to be a court expert His work in this area sullied what was probably otherwise a very good career Whether he was as he says 'trapped' into doing this work or not, he had agency, and he should have had the good sense to leave it to someone else better equipped to do it
So sorry for my late acknowledgement of your comment. Have just read it and I thought it was really good. It's great to hear your view on this and you make some excellent points.
Jenkins morphed from internal technical advisor for investigations, to internal Witness of Fact for prosecutions, to court Expert Witness but not independent crucially - incompetent guidance of legal codes to Gareth by POL internal lawyers, POL external lawyers, and judges. Lack of drilled down data analysis from Fujitsu, "tech snowing" by POL witnesses, and the unmerited respect of them by the courts, then roller coasted any defence who found it hard to contest or were ignored.
I came down on his side because two things. Firstly what he described in himself, I had actually seen in others when working. And secondly, he was not employed as in investigator of fraud and the like. So let me explain. Firstly those I have seen acting like him, They had tunnel vision and were focused on the task they had been given, and were not able to really look and assess what was happening outside of that. Because they were so busy, that to get everything done that was demanded of them they just had to plough on. And I saw this in him he was doing his day job, he was actually having to deal with problems that were arising as part of his real job then he was being asked to do this as well. Its more than obvious to me that nobody sat him down and told him exactly what the score was, there was no change or modification as to what he was doing at any time in relation to what he was producing. No attempt to change anything. You could clearly see where others were trying to make him say things and he was not prepared to say things. They said he was an unreliable witness because he had not disclosed things. But he had not been asked that question by anybody in the context of any of the trials or the situations. He had been asked to look at something in isolation by the post office. He was not specifically asked to look at the consequences of what he had found and if they had applied to other cases. Remember there were 900 prosecutions. He only provided evidence for 15 of them. Its the INVESTIGATORS at the PO that should have bene asking the questions that others that point the finger at him lay at his door And I saw a lot of that In Gareth. He was being fed a narrative by PO staff who were focused on the brand, and that is where his comments came from about the postmasters. He had never met them he was repeating what was being discussed. Secondly HE WAS NOT AN INVESTIGATOR. He was never asked to investigate the issues in the context of investigating a case of fraud. He was not asked to investigate what he had found could have caused in relation to criminal cases that he was not aware of. The prosecution team, the po solicitors and barristers, the PO fraud investigators.. That whole process, that whole team were not being properly managed and coordinated properly at all. I think that the system itself is robust but it does not have good checks and balances and the whole process of managing things with it was a sack of shitte.
I am so sorry for taking so long to acknowledge your comment. You make some excellent points here. Thank you so much for sharing your thoughts, you are very insightful and a nice bit of 'bugger it all' at the end that made me smile. p
Great presentation. The issue in Software Development is compartmentalisation, and third party vendors being part of the life cycle. If you break down the process into individual sections / modules and pressure is on regarding timelines / cost , it is inevitable for systems are rolled out unfinished. The only way to resolve these issues would be in-house monitoring teams that answer exclusively to the customer, and clauses in contracts enabling prosecution of developers . Scientism is the new religion, and people believe it can do no wrong, and this leads to shortcuts, that impact on people
I can assure you that whatever was going on at ICL, later Fujitsu, it was NOTHING like the Silicon Valley culture of 'move fast and break things'. This was a classic government/private IT contract and it failed exactly the way they all fail except in this case they managed to cover the failure up for a couple of decades. The British Government runs on fudge, or as they prefer to call it, 'creative ambiguity'. This is lethal when it intersects with computer systems which cannot tolerate ambiguity, and with competitive fixed price contracts which require a clearly understood scope. But the main culprit here is the legal system which somehow managed to convict people without the prosecution even demonstrating that any money was actually missing let alone that the accused had stolen it!
So sorry for my late reply John. Thanks for posting this. I think it's a good critique of my analysis and you do paint a picture of government that I think is better than the one I have. Nice work and thanks for posting this.
Maybe Gareth Jenkins was totally unaware of manipulative people and how they used him for their own ends. He was a degree carrying maths graduate from Cambridge steeped in code writing for computers. He did as he was asked by employers, too naive to see beyond the request. I think he got hung out to dry by everyone. A fall guy.
We have a relative in the IT industry and she often says that the IT software guys seem so focussed on the system itself that they often cannot see any implication of anything outside their circle of expertise .
This is a really interesting perspective. I agree that there is a moral code with software across B2C interfaces being used to pass the burden of corporate administration back to the customers. When there is a problem it becomes a user problem not a code design problem. Agents are left in the middle and are given a script to obfuscate "you are the only one experiencing this problem". The Fujitsu system was released early and I have always suspected, predicated on Fujitsu quarterly sales targets. I also think Jenkins wrote the code and tested it, but there appears to be a complete absence of additional scenario tests he may not have been responsible for - e.g. lost comms link, hardware failure and security architecture.
Agree with your analysis! He can be seen as perhaps an accidental, reluctant, and malleable expert witness. The most knowledgeable about the code, but not about the customer. He saw himself on POLs side - part of the team, and therefore had a similar level of curiosity as POL...
Hi Paul, just FYI, Microsoft will freely admit, going back to the early 90's that they released software when they could sell it, not when it was fit for use and they have had some real duds too like Windows 8.0 and Windows Vista. Back before windows server 2003, I earned a fortune in after hours call outs because Windows NT was so bad.
Thank you. I can normally make a judgment but with this I kept wavering. I think that he was naïve and believed what he was advised by information (or lack of) by Post Office, who were trying to cover it up.
Sorry for my late response. It was a tough one trying to figure where to sit when it came to Jenkins. I think your end judgement was a very fair one. Thanks for sharing :) p
I think your'e absolutely spot on , when you describe the users of Horizon, ie Subpostmasters, were more than likely and unkowingly being used as let's just say it ...'Guinea Pigs' ? as compares to your own analogy of Windows users etc back in the early 2000's to seek out all technical glitches ?. Seems to me that both POL & Fujitsu are equally to blame for the carnage that subsequently resulted .
I am so sorry for taking so long to acknowledge your comment. It was good to hear from you and yes, I think you are right about spm being the guinea pigs.
G" Day I have been following your comments on the Post Office and I like you originally a POM 1965 I came to Australia first Adelaide and now Brisbane I worked from the age of 43 until 73 with disability I had lots to do with your profession and found that they like Mr G Jenkins when trying to treat a person looked at the obvious behaviour and not the outer rim so to speak and had very closed attitudes so it's good for me to hear you without feeling that ha ha I am 80 yuck
Great to hear from you Lesley and thanks so much for posting. Good to know there's another POM nearby (though I am up in North Queensland!) Yes, that is one thing I do criticise psychology about too. They look at the obvious but then they put it through a meat grinder so it comes out as indecipherable mince meat and then they try to sell it back to us as scientific understanding. Well, that my warped view of things :) p
I have watched quite a few of the inquiry videos now and what sticks out to me is that everyone was at best lazy and incompetent and at worst ruthless and arrogant. echos of the 2008 financial crisis? of course the software didn't work as it should cos every piece of software is like that so to stand up and defend it for decades without properly investigating beggars belief (and I not convinced they could investigate properly either cos they didn't seem to have what should be a basic and fundamental mechanism to capture all the errors or report on them). I think people also have a natural tendency to rally around and defend blindly when things go wrong ! and you can still see that today in the defensive sitting positions of the Fujitsu and PO execs - and how strange is it they can't remember things if its about themselves but they can remember quite clearly if it related to someone else ! also Fujitsu would have been worried about not being paid for horizon work if there were problems so they weren't unbiased here and would have proactively buried their heads in the sand - its like asking someone to mark their own homework! from the post office side they were spending millions on these trials and no one thought to ask questions if this was the right approach - they would have been better off putting that effort and money into listening to the problems and fixing them - this speaks to a toxic ruthless culture. i have to say as well I am not sure about the value of the inquiry it just seems like a bit of theatre to appease the masses - cos nothing of consequence will happen
I am so sorry for taking so long to acknowledge your comment. thanks so much for posting. Yep, I do worry that this is just a piece of theatre with no action at the end. I really loved your analysis of the situation. Thanks so much for sharing. p
I doubt that the Horizon computer system is morally bad - its just a fancy networked shop till system The key decision was how to manage fault rectifications The system was buggy, and required huge expenditure to support it in-service When problems arose, PO needed to rectify them and support postmasters Instead they chose to blame postmasters, and prosecute them This was disastrous because the system could therefore not be rectified, because noone could admit the truth, and would inevitably blow up into this huge enquiry Jenkins unwisely got carried into it I think, rather than openly fixing software issues, as he should have been To me thats the moral bit of this
Some time ago there was an email (I think from PO general council) saying that Gareth had not been advised of duties as expert witness. Also that when he gave evidence at Misra he was not considered an expert witness (so wouldn’t have considered that note about it). There was a problem that the judge asked the expert witnesses to have a meeting and he was included. They then realised he had become an expert witness accidentally. I’m surprised this wasn’t pursued in this phase. The PO negotiated the price down so some of the visibility of tracing was dropped and information available from Fujitsu was restricted without extra payment. I think things would have been traceable (by Gareth) but the PO didn’t have an incentive to pay for that, Fujitsu didn’t want to do it for free and the PO concealed from the subpostmasters that it could be available. I think he was probably being honest - he was just interested in the system and was limited in what he was allowed to investigate. He should have been more obstinate about what he was willing to (not) do but had been with the company a long time and was probably concerned about leaving. I’ve been in a similar position and think it odd that he is meant to be in a senior (expert) position but doesn’t seem to have input into what work he does. I usually tell companies what needs to be done rather than the other way round - how can non-technical managers know? My opinion anyway. He’s just interested in the system and it was some else’s responsibility to use the information he provided. PO was just interested in info to support their position, Fujitsu were just interested in profit and keeping the PO happy cheaply.
I am so sorry for taking so long to acknowledge your comment. Really loved reading your thoughts on this and you make a lot of sense. Thank you so much for posting. p
Good talk. I do work with technology and can tell you have hit on a lot of truths about the changes in culture of testing. Regarding Gareth Jenkins, one of my roles is often dealing with the ego of developers and architects. These people are often detached sociopathic individuals. Have very one dimensional lives, they block out the reality they can make a mistake. Trouble with people at genius level they are incredibly blind to the idea they could make any mistake.
@blyndaza good question. I think not. My comment was primarily aimed at the comments by Dr. Paul about Gareth not seeming to care about the submasters' plight. I offered it as one possibility to explain Gareth Jenkins' apparent lack of emotion.
Hi Arthur, yes, I think a number of people would agree with you here. My take on this is that we are all different, some of our differences are exagerrated by the circumstances we find ourselves in and perhaps this is what is happening here. He did seem like a fish out of water. p
Placing moral responsibility, or lack of it in this case, onto the computer system is in in my opinion intentional. It’s a perfect way for management to step away from their own responsibilities. If anything goes wrong, if anyone is harmed then they can say it wasn’t me, it was the computer. In the case of our friend Gareth, I started his testimony thinking he was the PO’s useful idiot. But, as the days went by, I could see more intent. His lack of moral compass was being exposed. He was on message with PO lawyers. Still the usefully idiot but completely on board with trampling the little guy into the proverbial dirt.
So sorry for my very late reply. I am still catching up on the comments. Really interesting take on this even though I got squimish at the thought of trampling on the little guy. p
A great well-framed analysis Paul, well done. With the title of Distinguished Engineer, how could Mr Jenkins behave without considering moral consequences of the situation? Maybe he also failed to understand his obligations to the title? Moreover, I would posit that he failed to either abide by, or understand, his own institution's code of conduct? Just a thought.
In Mr Jenkins 4 day witness to the inquiry there were many instances and regular repetitions indicating neurological challenges. If this is the case the evidence may need to be filtered through an understanding of how this would affect all interactions, subordinations, articulations, vocal expressions, audible reactions etc. People all have limitations on how we gauge personalities due to our own experiences and analysis of others. Therefore, not all people may have the ability to recognise the signifiers or understand how it could affect and condition conduct with and around others and may lead to a subsequent misreading or misinterpretations of evidence given.
Technology has dehumanised ethical practice, governance & the simple trait of honesty. A rapid erosion of 'proper counsel & sound judgement' is a rarity in big corporations. Decisions are too often based on someone else's opinion of events instead of hard facts. Investigative processes only kick in when the damage has been well & truly done. It's all the wrong way round - our unhealthy reliance on the 'logic' of technology has seriously diluted good conscience. Love the brain-workout episodes you present. Keep going.
I am so sorry for taking so long to acknowledge your comment. You make some excellent points and glad that my videos give your brain a workout :) Thanks for the encouragement. It's much appreciated. p
As a former contract manager I have always hated the adversarial form of contracts, rather than a collaborative style. I know it’s very hard to achieve. It does not serve either party when severe and threatening clauses make it impossible to be honest with the buyer about issues. This clearly happened here, the PO screwed the price down so much that the operational side couldn’t afford the testing and monitoring involved. They couldn’t speak openly about issues and it encouraged them to lie about how the system was working to avoid penalties etc.
It would be interesting to get your take on the illusionary truths operating during the scandal and how various people in the different organisations involved responded to cognitive dissonance caused by evidence that challenged them. The obvious example is the assumed venality of the SPMs, but there were others, including the idea that Horizon was always operationally sound at a local level because it operated properly at scale. It feels like uncomfortable facts were repeatedly shunted into a “zone of irrelevance” through whatever mechanism was most naturally suited to the person involved, whether they were managers, technical experts, or part of the legal system. The degree of conscious agency that each had when doing this is something that the Inquiry seems very interested in.
Martin, I am so sorry for taking so long to acknowledge your comment. This would be such an interesting angle to take. I love the way you have put this - particularly love your term' zone of irrelevance'. If I can bring this into a future video, I most certainly will. p
POL were a knowing party to someone who passed himself off as an Expert Witness, when POL knew full well that, even after being issued relevant legal warnings, he persisted in passing himself off as an Expert Witness
Good analysis and interesting viewpoints. One aspect that you don't mention is that not all software is created or used equally. If Facebook fails and a comment on a post is lost, that is not a big deal. When Horizon failed, someone took their own life or was sent to prison. As for Jenkins, his explanation that he didn't understand his role as an expert witness is also BS, or does that apply to other parts of the law also, i.e. if you "dont't know" that killing someone us illegal, then murder is OK? As an IT system guy since more than 40 years, that is appaling, my morals are WAY higher than that. His excuses are 100% non-acceptable and he is a disgrace of an IT-pro.
Paul - software for 'safety critical systems' is subject to much more engineering rigour (i.e. standards, level and scope of testing, audits etc) as compared to the Horizon accounting software, as these can potentially result in the (more obvious) loss of human life. Accounting software isn't generally considered as 'safety critical' - however the actors involved in this case clearly brought about a loss of human life of another sort.
I'm a 1980s Computing Science graduate who's worked on many software projects, and I couldn't agree with you more about the 'morality' of IT development. Coders rarely realise that they're coding in their own assumptions, and these assumptions often come from a limited model of the world. Error-handling is critical; but developers frequently lack the imagination for all the things that could go wrong. My adult son is convinced that AI will be wonderful, and I argue endlessly that it's essentially just code, and that the code is written by (mainly) young men with narrow life experience, and that the built-in biases will colour the whole field. I've watched it happen again and again. Gareth Jenkins seemed to me just such a narrow-visioned technical wonk, absolutely sure of his rightness, and therefore very useful to the legal sharks who latched onto him.
Thanks so much for you post and sharing your experience. Really valuable to here this. I agree with your assessment of how Jenkins may have been used by the legal sharks. p
Oh this was brilliant! Thank you so much for your comprehensible explanation of technology related issues, especially concerning Horizon and players in the scandal! You know, if and when there are any prosecutions arising from the scandal I think your videos should be played to the Juries in Court! It would help them and others to understand what went so badly wrong! 🙋👍🙏⚖️🙏
I am so sorry for my late acknowledgement of your comment. thank you for your lovely encouragement. I really appreciate it :)
p
I worked on a very large government IT system that was 25 years old. I saw long serving IT people, who had been there so long that they were "untouchable". There are two power structures in this type of IT organisation. The official management structure, and then the IT staff who been around for decades. They have so much undocumented technical knowledge in their heads, that the management dare not mess with them. I think that Gareth Jenkins had a lot more power in the company, than it looks like on paper. I suspect that if management wanted anything done, it had to be done the Gareth Jenkins way, or it wouldn't get done at all.
Ironically, it was these sort of people who were pushed out the door when the inhouse PFI Contract bid was put together. As they had the highest salaries due to their very long length of service. So they were offered very generous early retirement packages. To reduce the total annual salary bill in the PFI contract bid.
That's how I view Jenkins too, taking on and nurturing the 'God of programming/IT' status.
@@ChrisM541 The problem is that they have to maintain their status as the supreme company IT font of all knowledge. So will never admit to ever being wrong. Or to not have an instant answer to any technical question.
I've encountered these rusted-on IT experts, too. I managed to get rid of the most obnoxious one a few weeks after taking over the team. That gave the others the clear message that no one was indispensable and a culture change was mandatory. Eventually the old system was replaced and the remaining 'gurus' were made redundant and pensioned off.
@@James_Bowie That's NOT quite what's being talked about here. You don't go blanket-culling(!!) experience like that! That's utterly shocking and totally damning on you and your management team.
When we finally heard from Gareth Jenkins, I thought he was going to be an awful person… but he wasn’t. He had his witness statements altered by others, he had a pretty one track mind - great in IT but knew nothing about law - possibly on the spectrum - but wanting to get things “right”- I put that in quotes because he was told his witness statements weren’t allowed to include mention of bugs errors and defects - so he thought ok - they want me to leave that out..,
The second sight guys had a breakthrough with their investigation because Gareth Jenkins told them all about the bugs errors and defects - because they asked him - so in a way, this whole house of cards came down because of his honesty - Ian warmington from second sight said he was one of the most honest people they talked to
Good point.
I am so sorry for taking so long to acknowledge your comment. So good that you have reminded us of this Doris. Yep, Second Sight thought he was one of the good guys and their judgement was pretty spot on for most of this.
p
Thankyou Paul for disecting the behaviour of these witnesses .
Jason Beer is disecting them with his sharp mind huge research obviously to come up with such focused questioning
A fabulous presentation! Erudite and informative on the wider issues outside the P.O. debacle.
Thank you so much for your lovely feedback.
p
I like your gentle, accessible approach to analysis of the Post Office / Horizon issues. Very helpful. Thank you.
That's lovely to hear. Thank you so much, feedback like yours keeps me motivated to make these videos.
Very enlightening video, I now have more understanding of why things went wrong with Horizon. Thanks
I am so sorry for my late acknowledgement of your comment. Just wanted to say thank you so much for your lovely encouragement. p
Thanks Paul - NZ here. Things I noticed: When NZ flight NZ901 crashed on Mt Erebus the airline was owned by the government and a cover up ensued. The public enquiry described the cover up as an "orchestrated litany of lies" to mislead investigators. The videos of the witnesses have similar "I don't remember" patterns as the PO Inquiry.
Government run/owned businesses seem to have a weaknesses in this area - your thoughts on why this might be would be interesting.
My background is software, network engineering, business management and also formal psychology study.
The contract between POL and ICL Pathway and later Fujitsu was for POL to provide the spec and Fujitsu coded - it's common in industry. Fujitsu didn't have suitable software so they developed one based on what POL asked for - likely Fujitsu had limited experience in this area.
POL could not provide a suitable spec as they were not competent to define a software system; postal mail companies are still very low tech, this was also part of where people failed. POL may have tried to replicate the paper based system they were familiar with - but paper systems have some inherent security and audit features.
POL used a contact that outsourced all risk to the SPMs, so there was no incentives to solve Horizon issues - they just charged everything back to the SPMs - so no losses and no questions.
Had POL brought in one of the big four accounting firms to work with Fujitsu, a better and auditable system that protected users (SPMS) could have been created. I suspect this is how govt control to keep spending down factored in.
The other angle is that all software has bugs, this includes all the banks and retail chains, but they manage the liability differently and so solve the issues rather than pursuing the staff to recover missing money errors.
Well said, not investing in the computer system at the start was where this all went wrong, from a technical level. To replicate the paper system was a big mistake, a completely new IT system, based on POL’s requirements should and could have been designed. However vast amounts would have been required in training the post masters to operate the new system.Similar to Boeing Max scenario, hide the faults and live with the crashes.
I recently watched a you tube presentation on the Erebus crash. Essentially someone somewhere sometime altered the on board co-ordinates for a possibly genuine reason but failed to tell the crew. This was covered up and everything blamed on pilot error..Very apposite
This is a good explanation of what has gone on since the ill fated useless Pathway software was tendered! 👍🙏⚖️🙏
That last paragraph is responsible for a lot of this.
Fantastic post. Thank you. Yes, government bodies do see to have a weakness here. I wonder if it is about authority structures combined with the value of duty - executive wield their power in the delusion they are serving the public and it becomes a kind of vocational zeal that results in real harm when they go off the rails. You can't impune their judgemetn because are a public servant and already morally impeachable. They think of themselves as self-sacrificing even though they are been very well recompensed for their work. Just some loose thoughts in response to your excellent question.
p
you make me think about the world more when I watch your vlog. So thank you.
Wow, thank you, very much appreciate that. And I am sorry for taking so long to reply!
Dr. DUCKET
Your calm, logical argument I find inspiring and something l am trying to instil into my own approach to discussions . You are one of a small number of contributers, I follow, who have the ability to make the listener think and arrive at the solutions themselves and not to take things at face value. I thank you for that! It may stop me from being drawn into heated arguments.
Thanks Raymond, that's lovely feedback to hear. I really appreciate it.
p
I've been looking forward to your analysis of Gareth Jenkins. Your assertion of the morality of the coders being present in the code is fascinating. I would extend it to include culture. I've worked with Russian, Indian, British and American coders, in particular. When I reflect on this, I have seen overwhelming evidence of cultural and individual by-products in their code, and these are nearly always highlighted when we've had to fix bugs.
I'm also interested in understating more on your thoughts of Gareth Jenkins himself. As someone diagnosed with Asperger's in later life, my subsequent research, reflection and understanding of the condition made me challenge my perception of Mr. Jenkins during his testimony. Whilst not to absolve him of any perceived ignorance or particular decisions, I hypothesised whether he had a flavour of autism which could have caused him to interpret, behave and act in certain ways throughout his work, and indeed, life. In some ways I abhor the use of the phrase "Common Sense" as I don't believe the term would exist if it were indeed "common". If we take for granted that Mr. Jenkins should have realised and thought differently then are we ignoring neuro-diversity? Society can't cope with trying to adapt to endless diversity, but could we better execute our work lives by structuring teams and assigning work in order to catch *some* of the more potentially impactful effects of neuro-diversity? In this case, I would propose that when legal teams and/or companies are preparing witnesses or employing technical architects and coders, that they apply more checks and balances on what will be the nature of the output of the people they are employing and how could they ensure a greater coverage of their needs rather than a narrow and individualised one. Unless, of course, they are choosing to use certain people in certain ways...
How do I feel about Gareth Jenkins after all this? I perceive that his contribution to where we are with this scandal was less deliberate and intentional than the contributions of many of the other witnesses we have seen. In some ways I almost felt an element of sympathy for him in that he couldn't comprehend how many of his actions had contributed to the scale of the issues. That's not to say that many of his answers or actions weren't prepared, cold, ignorant or selfish. On the scale of guilt, I see others as more culpable, and maybe even guilty of manipulating him and his character. However, he is responsible for understanding his responsibilities as a witness. Whether one is or is not Autistic, one cannot use it as an excuse in all circumstances.
One of the other things which, in my view, clouded his evidence was the possibility that he was, to a certain extent, possibly coached by his own legal team. I felt, at times, that he adopted a different tone. I wondered whether he did use some ideas which had been suggested to him by solicitors or barristers when speaking about some aspects of his work.
@@mikejames4540 Agreed. Some of his answers certainly suggested this. How any of the witnesses that repeated their "I cannot recall" answers can believe they would be seen as authentic is beyond me.
Great post Chris, you make some excellent points. As to whether Jenkins might be neurdiverse, in one sense we all are, just some organisations take account of that and removed barriers to our engagement and perhaps other organisations exploit that diversity in ways that aren't in the interest of the person or others around them.
p
Fujitsu turning a blind eye (with a view to minimising product liability) may be the crucial observation, because it leads to the question how Mr Jenkins was selected for his role of expert witness of fact (but, if possible, not an expert witness).
It is hard to avoid the thought that he combined deep knowledge of the Horizon software with almost complete personal blindness as to the meaning and consequences of the Horizon faults and of his own participation in prosecutions. He was the ideal intelligent idiot, so it would not be surprising to find he had been chosen just because of that. Moreover, as his evidence to the Inquiry unfolded, he turned out to be someone whose evidence in the prosecutions was easily deniable, in the sense that he never had first hand knowledge of the corporate processes surrounding the Fujitsu contract. And, if he could be offered as a witness of fact, he wouldn't have all those inconvenient extra obligations to the Court. So our sympathy for him is engaged, as it is for any other patsy. But his oblivious attitude then demolishes our sympathy (if not our credulity).
As for the morality of Fujitsu's approach to coding (and of the Post Office's attitude to procuring the system) this is, if anything, worse and more cynical than it has been portrayed. The system was not robust; it was fragile. It was expected to produce flawed output, probably for some time after it was rolled out, and consequently equipped with various ways of fudging the output from a safe remote location. It may have been hoped that bugs would be few and easy to spot, but there was a safety net for Fujitsu, that would operate at the expense of the postmasters. The moral issue for Fujitsu is that it was quite happy with this as a basis for delivering a turnkey system. It is not clear that Fujitsu expected the Post Office to prosecute the postmasters when Fujitsu fudged the branch accounts; probably Fujitsu was initially as surprised as anybody that this happened, once a week for years and years. When that happened, it turned to the oblivious Mr Jenkins.
For Mr Jenkins this was an unending nightmare, clearly. The only way he could cope with it was by denying the fragility of Horizon and thrusting out of his consciousness the impact of his brief career as an expert witness. The limitations of his memory were surely as much real as they were deliberate; he has had to spend 15 years preserving his own sanity. That does not excuse, in any degree, the way that he failed to take elementary steps to verify the sweeping assurances he gave about the robustness of the system, rarity of hardware failures and so on, when appearing before the Court. But that oblivious lack of care was what one might expect from such a carefully chosen stooge. The greater guilt lies elsewhere.
Well said Sir. Elegantly expressed, in a way I would have found challenging 🫡 .
Yes, if you look back at the evidence from Jason Coyne (v early expert tech witness, day 59, was expert in 2004)he highlighted issues, possible errors and discrepancies could be caused by Horizon. POL tried to get him to change his evidence, and when he wouldn't, got rid of him. You can imagine from that how pleased they were to find GJ, and how keen they were to continue to use him.
@@gailb8722 Absolutely! But I think it was Fujitsu that pulled the relevant strings to find Mr Jenkins, later on. The Post Office here was merely incompetent. Even Jarnail Singh seems not to have been making a clear-sighted choice, taking what Fujitsu gave him, when he was in a tight corner.
Yep.
This is such a good post. Really powerfully argued. Thank you so much. It is a really valuable contribution to the discussion and I think there are som many good arguments in your careful scripted post. Thank you so much :)
p
Having set a precedent, it is time that the Directorate, management and administration of the NHS is subject to a similar enquiry.
Exactly correct: having suffered at their useless hands for the last two years I can but agree, It is not more money that the NHS needs, but some solid management. Overstaffed, underemployed, overpaid. GP's are even worse: eight weeks to get an appointment, even though just discharged from hospital and needing further medicine - they are but a lazy joke.
Yes, I'd agree with that. Thanks Nick. Great post.
p
Wow! Super glad, I found your channel. I have been following the Public inquiry (also from Australia) partly because I lived in the UK from 1999 - 2007. I find the inquiry as a whole, fascinating and engrossing. I enjoyed your video and it has opened my eyes to yet more insights that I not even considered. Thank you.
I strongly believe this whole issue was a people and culture problem. Horizon is no different to any computer system - Jenkins is right about that - they all incur bugs and defects. It was the persistent culture of denial - primarily to protect the brand that is at issue.
True, but the Horizon system did seem a really bad system, many of the postmasters demanded to see their transaction records in order to prove their innocence, amazingly these records were/are only available for a month before deletion, contrast that with say Amazon, where every purchase you've ever made can be called up instantly, or your banking records, ditto, the whole post office culture was to preserve the facade of software integrity at all costs to please the government and shareholders.
The problem jenkins said the postmaster very filthy crooks
@@monteceitomoocher 'many of the postmasters demanded to see their transaction records in order to prove their innocence'. This is the crux of the matter: we have heard again and again that the postmasters 'offered no evidence' that the Horizon system was faulty. How could they, they had no access! Besides in British law you are supposed to be innocent until proven guilty - somehow the burden of proof got reversed in these cases and that is a defect of the legal system. Astonishingly cases were allowed to proceed all the way to conviction without the Post Office even having to prove that any money was actually missing let alone that the accused had stolen it!
@@john_hind Yes. But the position of the P.O. counsel is far more culpable than you have indicated. I am informed that the Fujitsu equipment -- (either at head quarters or elsewhere) -- had a (permanent?) record of EVERY KEYSTROKE entered by a SPM. This record would provide CONCLUSIVE evidence of any misappropriation of funds. Why was this evidence not made available -- in the case of a criminal trial? Who made the decision to withhold that evidence?
Yes, I agree it's really important to keep a focus on the people and not get distracted by the tech.
Thanks for that.
p
I was at the Inquiry for his evidence from Wednesday to Friday. It was fascinating. I agree he was a hard one to work out. I had a lovely chat with a lady who was also attending about what we thought of various witnesses and your videos came up as we'd both seen and enjoyed them. Just thought you might get a kick out of knowing your videos were being discussed at the Inquiry 😂
Well, you're right. I got a tremendous kick out of that. Thank you so much for letting me know. That has really given me a boost. And, well done for going to the Inquiry. Having people like you there keeps them all accountable :)
p
So I used a computer system designed in 1989 in a major uk utility firm. Everything i did had a used ID and time stamp.
Even if I looked up a customer or an address it was recorded..
After a shooting , a number of curious people looked up the shooter, and got reprimanded.
The idea that you could change a balance without a data stamp and user ID is madness…
Even if the system changed a balance. The system should assign its on ID.
????
Totally concur and that balance transaction date/time/User ID stamp would have another date/time/User ID stamp against it when it was "corrected/adjusted" by a Fujitsu "adjuster" showing it had been, in these situations, been "tampered" with.
That is such a good point. It's kind of the thing computers are good at doing. Thanks so much for posting this.
p
Yes, but you were an END USER.... just like the spms were. *Fujitsu own and have access to the CODE fgs!*
Of course they can do this. It's a no-brainer.
Single entry adjustments were made by Fujitsu engineers directly into the backend.
What was missing was the appropriate security/governance/signoff processes and procedures.
When Jenkins sent his witness statement to Jarnail Singh, he also sent a memo in attachment.
In it, he described three possible solutions for branches affected by the receipts and payments mismatch bug.
TWO of the three options involve doing PRECISELY this.
*As he described it on the second day of his appearance at the Inquiry: 'it would involve changing Branch data without informing the branch'*
The memo also speaks to the RISK of each option and the moral implications 'to the Integrity of the business' of manipulating data in this manner.
THATS part of the problem with Mr J.
Not only did he suggest and describe this in that memo - which we clearly saw in Mr Singh's testimony AND Mr J's - the memo was an attachment to the same mail he sent Singh his witness statement, in which he said:
"No external systems can manipulate branch accounts without user awareness and authorisation."
-- Gareth Jenkins witness statement 2010, for the Seema Misra trial.
*He knew this was not true.*
Thought provoking video - I have always maintained that any business should be an honest endeavour based on fair dealing and benefit to all concerned. Larger corporations seem to become arrogant and Machiavellan over the years treating many of their staff. customers, suppliers etc as useful but disposable but the PO scandal seems to have shone a light on this assertion. But the question is why do so many at PO seem blind to their lack of deeper insight or their own morality ?
Sorry for my late response. Really good points you make here. Yes, I thnk you have identified a key question. thanks so much for posting.
p
Appreciate the balanced view , perhaps if more people like your goodself had of been involved we just might not have been in the position we are in , keep up the good work
That's lovely feedback. Thank you
p
In your conclusion, you were able to reduce the scandal to the essential problems, nicely done !
Thank you so much. So please it came together at the end in a way that worked for you. That's great feedback to receive :)
p
I found it interesting that Jenkins was still very defensive about Horizon. But most notably to me, he couldn't adequately explain why he had repeatedly failed to reveal important information that would have pointed to defects in the system.
Great observation. It's often key to look for those contradictions and you brought an important one to the surface. Thank you.
p
It was his baby after all - but I think he had been admonished after a very early court case by POL or Fujitsu to not reveal Horizon inadequacies so openly in court again, so being a company man, he did not..
Dear Dr Paul…
The thing that struck me about Gareth Jenkins’ testimony was about the standard Fujitsu witness statement by himself and Penny Thomas and Andy Dunks that the computer was “working correctly”. Gareth pointed out this wasn’t the Horizon system/computer but was the computer used by them to extract the audit data. I thought that was a significant point that I had not appreciated or considered previously, and I don’t think the Inquiry had considered either.
That was the one thing he said which I thought was a lie to cover up what it obviously referred to.
@@cassandratq9301 Thanks for the replies. Gareth said that Penny Thomas and Andy Dunks would have no knowledge as to whether Horizon was working correctly, but that the 'computer working ok' in their prosecution witness statements was a reference only to the computer used to extract the ARQ Data from the audit data base. Andy Dunks is giving further testimony today - so it will be interesting to see what Andy has to say on this.
This line of questioning to a number of witnesses both confused and irritated me.
As I see it the original clause (drafted by lawyers in both Fujitsu and Pol) referred to the service by which Fujitsu extracted transaction data from the archive and presented it to Pol, and was meant to certify the validity of the data and the function of the extraction mechanism… “the computer”.
later Pol was in the habit of drafting witness statements for Fujitsu and used the same template, the clause had no place in a witness statement which did not include transaction data extraction but was still included. Hence Jenkins’ assertion that it must refer to the pc that he typed the statement on. The inclusion probably suited Pol because it was unclear.
Later still, by persons unknown, “the computer” became “the system” which gets Mr Dunks and others into a world of trouble and suits Pol even better.
I don’t believe that the inquiry does not understand the provenance and timeline of the witness statement pro forma and it irritates me that they use its confusion to attack witnesses when they should be searching for the truth.
My take on Jenkins…. A very busy tech “beard” trying to be helpful and trusting the legal system to look for the truth, I take you to the testimony of the prosecuting barrister In the Castleton case - “it’s an adversarial system and I won”
Sorry for my late response. You make a great point. Thanks so much for sharing this with us, it was a very good spot.
p
thank you very informative. To me there are two angles to this scandal which are related. Firstly the technical lack of moral code which is highlighted in the video. Secondly the commercial, legal and career preservation lack of moral code(s) displayed by the PO management at all levels.
Great point Jonathan. I think you are right, the commercial and legal factors are key here.
p
“Oh, just one more thing,”said Lt Columbo. GJ worked his entire life at ICL (ignore the name change to Fujitsu for now). By the time it got the post office/benefits agency gig in 1997/8, it was a busted flush. Its only customer of note was UKHMG. It was British (huzzah!) and it was cheap. Its technology was poor. In many cases ways it resembled British car production, traded on former glories. Just thought that’s worth remembering in all this. I can’t wait for the ‘Horizon’ movie!
hahaha, nice one :) I think tech companies have a habit of dumping bad code on governments and the like. I see it all the time in education - until Google and Microsoft woke up to how schools and colleges had children who would grow up to be tech users and so needed to get them young and get them on brand, techn companies used to just give schools crapware because they weren't seen as 'enterprise/business' and usually didn't complain.
@@PaulDuckett Thanks Paul, good insight. British exceptionalism and our fetish for flag carriers doesn’t help. I worked for an offshoot of Cable&Wireless, aka the telecoms branch of the Foreign Office. Boy, did the empire live on there. The C&W chairman had a butler. The Mercury building in Bracknell was opposite ICL hq (early 1990s), car park full of range rovers, range rovers full of golf clubs. They were already crap, then Wang and Dell arrived, almost like mourners!
Thanks, very interesting. For me, the "high level" analysis would also include alienation - the alienation that results from an economic system that prioritises financial profit over any other socia/human outcome of activity, and the alienation of individual workers located within a corporation so large that they have no contact with the people their work affects
really good point. Thanks so much for posting
p
Thanks for your thoughts. The moral ingredient of computer coding is surely one of the least discussed and yet possibly one of the most important.
Your questions are very salient and quite disturbing. Thanks again for your calm and well-articulated analysis.
Thanks for this. Insightful and thought provoking. A healthy work culture and climate is vital to protect us all when things go wrong.
Sorry for taking so long to respond.
Yes, I absolutely agree with you on this one.
p
How interesting about computers and moral code. We spent a couple of days at the Public Inquiry and came away with a very similar feeling to yours about Gareth Jenkins.
Sorry for my late response. Great to know you are attendin the hearings - it's so important that the public fill the public gallery. Great to know we saw this in the same way.
p
Dr Paul, thank you for your analysis. Tying the Horizon scandal to the "Move fast and break things" mantra of Facebook and the wider IT industry is an interesting thought, particularly when recognising that the mantra had its genesis in software development practices such as Extreme Programing and Scrum methodology, which both date from 1993.
To give him his due, Mr Jenkins was not the design lead architect, but seems to have inherited the title of "Mr Horizon" some-time around 2003. Too many of my colleagues fail to make the intellectual leap, that software is not morally neutral, and that we have a duty to Society to do all that is humanly possible to produce software (and attendant manual systems) that act as good moral agents.
I am so sorry for taking so long to acknowledge your comment. Thanks so much for sharing your thoughts and I think you are right to give Mr Jenkins his due that he didn't design that thing, he just got sucked into it all.
p
As a former contract manager I have always hated the adversarial form of contracts, rather than a collaborative style. I know it’s very hard to achieve. It does not serve either party when severe and threatening clauses make it impossible to be honest with the buyer about issues. This clearly happened here, the PO screwed the price down so much that the operational side couldn’t afford the testing and monitoring involved. They couldn’t speak openly about issues and it encouraged them to lie about how the system was working.
@allisonstandley2573 Jenkins would say he didn't lie, or at least didn't mean to.
He appears to have been carried along into agreeing with statements that PO wanted, and he was unwilling or unable to prevent them
Whereas PO seem to be institutionally corrupt, Fujitsu seem to be inept and unable to define a robust position for themselves
Yes, I agree. It's those corporate contracts that can be so horid. PO was supposed to have a relational contract (one premised on negotiation and discussion), but they didn't do that which is what caused such horrid outcomes.
p
I agree. A significant reason that I left Fujitsu in 2001 was that I hated this kind of project, which accounted for an increasing part of FJ business. You end up with customer and supplier constantly trying to score points off each other rather than working together to a common purpose, which I have seen happen to great effect in private sector projects of similar scale. It's inevitable with this kind of large project that the last few bugs will be found during live running, and that isn't a problem if everyone recognizes that and works together to deal with the fact. Openness with the user community - sharing information with them and encouraging them to share information with each other - is a vital part of that, and is what was totally missing on this project.
Very good point about morality and technology. Thank you
Thanks so much :)
p
It was interesting when Mr Jenkins said that he was better with computers than people. I have the feeling that would apply to many people drawn to working in IT. At times he appeared utterly bewildered and unable to understand why people were so angry and frustrated by this situation.
Sorry for taking so long to respond. Yes, I agree, that was interesting when he said that and does capture his predicament, I think.
p
Thank you Paul, very much appreciated.
It's my pleasure :) glad the video worked for you.
Hi Paul I think you work on this is amazing please keep it up. It’s helping me understand the mindset of the post office.
Thanks Dr Paul, very interesting. I was struck by the way that in his reports to the court, GJ answered very narrowly the questions that he was asked. In his oral evidence he said he hadn't considered the wider picture, that no one had asked him, and I believe him.
If you look back at the evidence from Jason Coyne (v early expert tech witness), he highlighted issues, possible errors and discrepancies could be caused by Horizon. POL tried to get him to change his evidence, and when he wouldn't, got rid of him. You can imagine from that how pleased they were to find GJ, and how keen they were to continue to use him.
The way GJ struggled with open questions, and answered tech questions with rapid very detailed enthusiastic answers seemed like something I've seen in many software engineers and similar. As GJ said, "Better with systems than with people".
Thanks for your interesting analyses!
Did Coyne's fate influence Jenkin's behavior?
@@cassandratq9301 Interesting question! I would say yes, and Post Offices's behaviour of keeping using him and not risking finding another independent expert!
Sorry for my late response. really good reflections. I didn't spot the narrow answers, but I think you are right. Thanks for sharing your analysis. Really worth reading.
p
Horizon was shit but wasn't the problem for the SPMs. The problem was the ignorance and fanatical nature of POL corporate governance and the pervasiveness of some of hateful social attitudes.
Class attitudes.... we are dispensable..
There were two levels to blame. The middle management with Horizon and the top level with hateful social attitudes.
Nicely put, Thanks Matthew,
p
If horizon had been perfect then the issue would not have happened at all, but of course no software is perfect.
If it had been good enough then they would have had occasional issues that they could investigate. This didn't happen either and the prejudice was allowed to take over
You can't prevent prejudice, so I would argue that horizon was the main issue.
@@phill6859 I very much think you will find that the prejudice element existed long before any Horizon issues.
First, I really love your channel, your content is always thoughtful, thought-provoking and informative. Second, your take on Gareth Jenkins chimed with my own. He seemed to see things in quite simplistic terms and I even wondered at times whether he could be on the low end of the spectrum. He looked through documents by ignoring anything that he thought didn't apply to him. He did what was asked of him. It's evident he was vulnerable to being used by people who had an agenda to push. This doesn't, however, remove his culpability and his apparent inability to appreciate the human cost. Maybe you could say he knew how technology worked, not humans. Thirdly, to me this video didn't seem too long at all. In fact when you said a couple of times that it was coming towards the end I felt quite dismayed, and said out loud "no! I could listen to this all day". Thank you. Look forward to the next.
Thanks so much for your lovely feedback and your thoughtful post. Yes, the question of Jenkins' neurodiversity is an interesting one. I think we are all neurodiverse in one way or another but that some forms of diversity are exploited by companies like Fujitsu and POL to cause harm. I think that Jenkins' might have been exploited in that way. Perhaps that is why we couldn't blame him as much as we thought we could.
p
- your comments made me think about the ethics approval I needed for my qualitative research and you're right, once ethics approval was granted I implicitly understood my research to be ethical. Perhaps even moral to some degree.
Thus i paid it no further heed.
I don't know of the any ethics approval or overseeing body which regulated the Horizon system.
Shirley future court cases will focus on this failing.
All of the wretched individuals in this sorry episode and injustice, to my mind, have essentially minimised their individual involvement and culpability.
😕🥺😕
Excellent point James. Yes, ethics has to be continual rather than episodic. I think a lot of folk get caught out by that. Sorry for taking so long to respond, your post was excellent, I'm just very tardy!
Fascinating and thoughtful in equal measure.I spoke to GJ's KC prior to his evidence and also Fujitsu(they are not representing him) and suggested that he was in my opinion incurious about the detail of matters outside his work/interest,typical of Senior IT people..She replied "lets wait and see2.The question for Fujitsu is"How many other Dept. did you have that supplied prosecution support to bring end users to court?
I just want to know it was determined which postmasters accounts were corrupted and how the amounts were determined.... If it is true, that the postmaster who actually saw the staff was then targeted to the tune of 40 some thousand pounds, there is a record of that in correspondence email or paper or bank accounts... The code allows the data manipulation/insertion. but somebody is telling them which accounts and how much....
Thanks so much. I think what you suggested to his KC was a really useful question. Nice work, well done. :)
p
Love your insights and the way you deliver them. No high drama just the facts
So sorry for replying so late. Glad I got to your post, it gave me boost. Thank you.
p
This 3 day interrogation of JG had many commenting on Live Chat perplexed too. It has been suggested several times and on different days that GJ is on the autism spectrum. The reasoning being - it was suggested - that this would account for the detached way in which he spoke of events (note: not in an ‘aloof’ way but ‘detached’ and ‘coldly objective’). It was also noticeable that he liked the technical questions from the change in his general body language and demeanour to his interrogator. Concrete questions good, value judgement questions bad. This meant (it was further suggested) that he could be taken advantage of and by the lawyers in particular as he judged these important people by his own standards.
Exactly.
So sorry for my very late reply. I am still catching up on the comments. You make some really good point. I didn't realise that was happening in the Live Chat. Gosh, very interesting!
How many of those responsible will face a criminal trial and prison? The Post Office inquiry is theatre to appease the masses. We saw with Partygate that laws are for the little people.
Correct no one is going to jail
100% agreed, I’ve said this all along! It’s obvious, inevitable and typical!
And very expensive theatre it is. Done better to get back all the bonuses, not spend millions into barristers pockets with this droning on enquiry and pay the Sub Postmasters substantial amounts that is tax free and their legal fees paid by the post office.
@@brianwithers162 agreed 👍🏼
Yes, sadly I share your pessismism on this. Thing have gotta change.
p
As someone who has worked in the nuclear safety business for more than 25 years, I have always believed that an ethical dimension is an essential element in our work and MUST be explicitly identified. Unfortunately, like in the post office, this has been treated as “irrelevant” to what is treated as a simple technical issue. Thank you for your thoughtful and very accessible broadcasts.
Ethics are not included in your employment contract, in fact they are abrogated by any NDA you might sign.... Imagine that... You think you will get a bonus for your ethics but that is purely delusional in the current culture...
I am so sorry for my late acknowledgement of your post. Great posting, thank you, and thank you for you lovely encouragement. p
Paul, I think that while your argument about the morality of code is thought provoking and a valid one it is a stretch to apply it to the Post Office Scandal/Horizon. It’s a system working with financial transactions - the scope for morality in the coding is very limited (if it’s there at all). I can only see that it creeps in in a very human and insidious way in terms of the coding allowing access without notice to the sub postmasters. The overriding point has to be the human (inhumane) reaction or inaction to the flaws that were there.
I have to disagree. My version of properly moral code would have had the back-door inserted transactions clearly identify who made the (ahem) "correcting" transaction, at what time & day, and from where. The Horizon system had none-of this. ergo: a purely technical financial, yet morally compromised solution.
Thanks Goffers, you make a good point. I may have over extended my argument. But glad you engaged with it nonetheless and posted your thoughts. I really appreciate that.
p
This was very interesting and thought provoking. Thank you
I am so sorry for my late acknowledgement of your comment. Thank you so much for your lovely encouragement. p
Thank you for such an objective analysis. The UK loss is obviously Australia's gain.
I am so sorry for taking so long to acknowledge your comment. And it is such a lovely comment too. It gave me a lift :)
p
You dont understand tech, that isnt what is happening and cant be done in financial systems in the UK . Yes all software has faults but this company is unregulated unlike every other financial business. They prosecuted people for stealing and didnt return the money to where it was suposedly stolen from. Gareth lied in court repeatedly and then after the PO were told not to use him anymore they still did and he co-operated throughout. They found someone who was prepared to keep saying what they wanted in court despite evidence to the contrary. The software broke ACID rules, he gave evidence on the quality of the code but couldnt answer Mr Beers question if there were unit tests in it. Not that would have saved the errors but it proves he hasnt looked at the code.
Didn’t return the money? Eh?
Thanks for posting. You are right, tech is not my area of expertise. It's good to hear your different take on this. Thanks so much for sharing as you've made some good points,
p
@@annoyingbstard9407 no they added it to the post office profits....yea really
Thank you for a very enlightening insight into computer coding and it’s consequences.
I think there are several factors that might explain Gareth Jenkins contribution to the Horizon scandal. First, he appears to be what we now call a nerd or a geek. He is obviously an accomplished IT engineer - hence the Distinguished Engineer title. People who are very good with numbers (mathematicians, statisticians, accountants, data scientists) are sometimes lacking in another critical area - words. These people often feel more comfortable processing numbers (calculations, equations, coding, etc) than talking to others or expressing ideas or opinions unrelated to their specialisation. To his credit, Jenkins answered all questions put to him in the inquiry patiently, clearly and (surprisingly) concisely. He considered every question carefully and gave accurate and consistent answers. It is clear why Fujitsu chose him to be the expert witness.
The second factor that may help in understanding his behaviour is loyalty. Jenkins had essentially worked for the same organisation his whole professional life. To use a hackneyed phrase, he was a company man. He owed them.
The third factor is pride. He had the Distinguished Engineer title bestowed upon him by Fujitsu. To criticise Fujitsu or Horizon is to criticise himself. Also, he was instrumental in designing the Horizon system - which he continued to defend throughout the inquiry. To accept that the system was faulty meant admitting that he had failed in his role or had been negligent. Neither was palatable. From his standpoint, the system worked as it should. BEDs were at an acceptably low level. The only problem was that the consequences of those BEDs were convictions of real people. And "people" was an unfamiliar concept to Jenkins.
Well said.
@@cassandratq9301 Having said what I said, I believe he does bear responsibility for the evidence he provided in the trials of SPMs. He must have understood that the evidence he gave would likely result in successful prosecutions.
I am so sorry for my late acknowledgement of your comment. Really great response Cassandra. I think your analysis is really good on this.
p
It seems that the public and various bureaucracies have outsourced their morals and ethics and so they say, ' Not my decision, idea or whatever'; not my fault'!
Worked for Raygun Ronnie who tripled the US Debt..
Very nicely put Lynne,
thanks :)
p
I do like your reasoning Mr. Duckett. Steel sharpens steel is a tenet we should all live by in debate.
Nice! Thanks for your succint and sharp post :)
p
I'm loving your take on things.
The morality of coding seems so obvious they way you present it, but yet it was absent in so many people's thinking.
Maybe because it can't be measured and there's no expectation for accountability?
Me? I'm just sick n tired of inadequate corporate processes. It seems everything is designed to go wrong, and does so often.
Hi again from NZ Paul, thanks for your video and if you hadn't mentioned yr own laptop issue, I wouldn't have noticed - all the same you are very clever with your editing tools.
You raise very good points here about the process of software (SW) design, development, testing and eventual release to end users and its potential impact of unintended consequences (the social/moral aspect).
My SW background is mostly in military and civil aviation and the consequences of getting it wrong can be catastrophic of which the Boeing 737MAX is sadly an example of.
In the case of Gareth Jenkins, and no disrespect to SW engineers/coders is intended here, he reminds me of a very small number of these folks I've worked with who think only at the bit and byte level, ie down in the weeds.
When given a well defined spec with clear expected outcomes, most will meet the desired goals. However if the spec is poorly defined then the SW will most likely not achieve its desired outcomes and that is clearly where Horizon sits I suspect.
This said, Gareth, like the others in the management chain here need to be made an example of as with what we've seen to date from the enquiry, the comparison I'd make is they've all been driving at speed in a car that has been running on only the wheel rims and they've chosen to ignore it simply by turning up the car stereo to drown out the noise in the hope it will be alright and go away.
Thanks again for your videos, I find them entertainingly enlightening.
Sorry for taking so long to respond Brent. Just catching up with things after being away from youtube. You make some really important points and it's great to hear your views on this given your tech experience. Thanks so much for sharing. And, thank you for your lovely feedback too :)
p
Watched hours of testimony from many of the witnesses, often make a decision on what I think they did. Then Mr Beer moves in a direction that goes against my thoughts. What I learned is that without all the information, we are likely to make the wrong decision on guilt. Only view I would trust is the final view at the end of the inquiry. No stone left unturned
Couldn’t agree more, it does show that the enquiry team work incredibly hard to winkle out the truth.
Great post. Thank you.
p
My husband has just watched this video. He simply commented that Dr Duckett reminds him of an old sociology lecturer of his in the 1980s, Dr (later Prof) John Hughes at Lancaster University in the UK.😊
Sorry for my late response. That's so funny. Hopefully your husband thought Prof John Hughes was a nice fellow :)
Another fantastic analysis, I worked in tech for 38 years , and was beset with "B.E.D's" and the end user had to find these and we as field engineers reported them back, customers lost faith and took their business elsewhere but ultimately found the same issues, so you're absolutely correct it is endemic in the tech industry. Ultimately GJ wrote that code without careful consideration to the consequences of his failure to write it properly therefore he IS responsible for those consequences but failed to acknowledge that . But it's not a PlayStation game that you just press reset and start again. No ...its immediate custody and ruin of innocent people's lives . Thankyou for these brilliant videos Dr Duckett 😊
There is a crisis in Academia , with thesis / studies using faked results to reached the intended goal . There was an old idea of a " No Blame " culture, this allowed admission of mistakes .... we need to revisit that idea
No quality control then?
@@Andrew-rc3vh basically...no
Thanks so much Mark and great to hear your insider experience on this.
p
Rather a simplification, I think. Firstly, I doubt that GJ wrote much of the code himself. In fact, I doubt that FJ wrote much of it: a lot of the problems were that they selected the third-party Riposte sofware as a key central component of the system, and it proved insufficiently scaleable for the transaction volumes being handled here. The software did eventually get to the point where it was reliable enough to meet the operational needs of the business, the problem was that that's not the same as being reliable enough to be used as the sole evidence to prove that users have committed a crime.
I felt the same conflict. He came across as quite lame and was clearly filling a role he neither chose to understand nor was qualified to fill. Seems willful in his ignorance as he was enjoying himself too much. He didn't illicit anger (as some have, deservedly so) but he is definitely culpable as a willing participant.
Great thought provoking video, as always. Thanks.
So sorry for my very late reply Stewart. I am still catching up on the comments. Really appreciate your comment and yoru encouragment. Thanks so much. p
I think one important distinction in the Facial Recognition comparison is that Horizon was suffering bugs while the Recognition software was functioning as they'd designed it (but that design was flawed). when Horizon added things to balances when it shouldn't be, it was performing actions that shouldn't be part of the process. the FR software is, for want of a better term, "working as intended" but it's just bad at its job. the racism coded into it means it does poorly on black faces, but that just means it spits out the wrong score/result for its comparison check. it's a flaw, but it's not causing unintended behaviour; by which I mean it's not like when the software sees a black person then it suddenly starts trying to print the picture or ask for the user's login details. it performs poorly, but it still performs the actions it was told to perform.
Horizon behaved in unexpected ways. it initiated processes and did things it shouldn't have been doing. so while the comparison is interesting (and I think has some merit in terms of thinking about the impact of code and being careful when you write it), I'm not sure if it's the most applicable
I think Jenkins genuinely thought of himself as a 'God of IT/Programming', and as such, took the stance that Gods take instruction from no one. I genuinely think he dismissed so much because he though it was beneath him/bothersome to engage. He even took on the title of 'Dr' with no objection...or correction. However, you do find these characters particularly in computer programming, a field where the socially awkward can safely gravitate. What he doesn't seem to grasp is the significance of software bugs/defects on the 'users'. I bet the film Tron sits high in his favourites list.
He was most definitely used to the maximum by the PO's legal team, clearly 'allowing' him to leave out crucial evidence, or rather, informally instructing him not to 'disclose'.
Great post. Thankfully at my place of work our IT folk are fabulous. They don't they they are gods but the rest of us do. But I completely recognise what you say. It's when knowledge is used a authority. That can be really risk as knowledge is often flawed an incomplete.
p
I, too, am conflicted in my opinion of GJ.
Gareth Jenkins was one of 30 developers at Fujitsu, and 30 shared service centre staff. There were many layers of Fujitsu above him, and many layers of Post Office Employees above Fujitsu. Whilst it's right that GJ's part in this miscarriage of Justice is scrutinised, he is a small fish in a big pond, and it would be wrong if he was the only one so heavily scrutinized.
He was remarkably blasé about his evidence provided to POL and how crucial it was in convicting many sub postmasters. This suited POL and Fujitsu, at the time, and now.
His part in providing evidence seemed to evolve over time (boiling the frog analogy) , and he wrongly, never seemed to grasp that he was an expert witness with the same responsibilities as an external expert witness, and wasn’t just providing a couple of paragraphs for a corporate report to be used by a customer of Fujitsu.
In his early evidence, he stated that Horizon was 5%- 10% of his time , so he had many competing objectives for his time. He provided circa 15 witness statements over a dozen years. His reports, commissioned for different reasons , were used many more times, both in court and by Senior POL management, but he was unaware of this until the Inquiry
He was remarkably casual in providing blanket assurances, without making an effort to research if they were actually true at the time, and continued to be true, or just assumed to be true, based on informal office conversations, or that had been true when he first learned about them.
To determine if these assertions were true , would have taken significant additional resource, and he could have refused to provide those assurances, without the extra time and resource. He didn't. POL solicitors, having received those assurances of certainty, from an expert witness, were happy to take them and use them to prosecute.
It's clear now that the functionality, to insert transactions at Post office Branch level, was specified and built from in the start.(2003) It may have been intended (rightly) to be used rarely, with high levels of authorization, clear transparency and auditability. As it was actually deployed, none of these were true. Many people could use it, inserted transactions, were not authorized by Sub postmasters, and the identity of the users who posted the transaction could not be relied on to be absolute.The audit reports available did not show the crucial information, i.e. externally inserted transactions.
The use of inserting transaction at central level, by many people , had become common place at the shared service centre, possibly because of faults in the software, and the need to fix them to allow business to continue at Post office branch level
Externally posted transactions were often identified as the last local user logged on (usually the sub postmaster), not the shared service centre agent or developer. The actual user could only be identified by extra work and analysis of base level transactions, none of which the sub postmaster had access to. The standard reports showed the "provided" user ID. Investigators rarely asked for the extra analysis, because they never knew if the User ID on the report was a problem. There was no incentive for them to ask for this, as it took time and money, and would only weaken any prosecution case. They received cash bonuses on successfull prosecutions.
The sub post master was under intense pressure every day, to balance the books correctly or accept their culpability for the out of balance amount, or the branch could not be opened the next morning. It appears to be common place for sub postmaster s to be in the office until the early hours of the morning, and then having to put through a balancing line, or not open at 9:00 a.m. “Caught between a rock and hard place” for a small business owner. Imagine having to leave the shutters down and explain to the queue, "I can't open because I can't balance my books"
When originally deployed , the out of balance amount could be posted to a suspense account. But the amounts on these accounts were often several thousands pounds. This was a worry, so instead of finding out what problem was causing the need for balancing amounts, they blocked the sub postmasters from posting a balancing item to the suspense account. This accelerated the values of unexplained out of balance amounts, and potential prosecutions
The wider issue of the unequal contract between POL and the sub postmasters, encouraged POL solicitors to prosecute Sub postmasters to recover "missing funds", rather than pursue the root cause for the errors, which were more than likely problems with the Horizon Software, and therefore for Fujitsu to fix.
The lack of scrutiny from POL to Fujitsu on the integrity of the Horizon system, suited Fujitsu commercially, as they were engaged in selling additional contracts to POL and other Government agencies throughout this period, so having a key platform branded as unreliable would have made winning new business almost impossible. So GJ as a useful idiot suited them
Horizon worked well for most branches most of the time, but there were regular issues. These issues seemed to arise with rare combinations of events and were often difficult, time-consuming to replicate, and fix, and new ones continued to appear over time, with new combinations of hardware and software and operations. IMO this is pretty normal, especially for bespoke computer systems. Software company fixes them and life moves on. It doesn't usually end in someone going to prison
Fantastic summary. (I've been watching the Inquiry from the States.)
So sorry for my late reply. I think you post was really important . Thanks so much for taking the time to lay out the issues in this way. It's very valuable and I am very grateful. I think other's will benefit from the careful thought you put into this.
p
@@PaulDuckett No problem GJ' was painted as a villain, by the media, so I was keen to understand how much he was to blame. After watching several days, it was clear to me, a newcomer, that it was more than a one person issue.
@@Pat14922 Indeed, it was more than a "one person issue". However - one critical aspect of GJ's testimony fails the sniff test to me. I find it incomprehensible that someone with the intellect to achieve a Maths degree from Oxbridge can claim not to understand or known (or research) the difference between a civil and a criminal case, especially if knowingly providing a statement (let alone an expert witness statement). On repeated occasions. Including appearing in-person in one of the seminal cases (the Misra case). To me, that's a line he was coached to take. There's clearly been massive coaching taking place for inquiry witnesses, to the point that much of it is useless.
I speak as one who's done the IT expert witness bit just once in the late 80's, in a civil case. But then I'm a mere Imperial College engineering grad....
@@mpellatt Massive coaching? Agreed.
It's written all over their testimony. The witnesses know a solicitor will be poring over their testimony on this inquiry in deciding who will face criminal charges.
GJ - is quite likely to face a trial, and the court will then determine if he was ignorant or criminal
My concern is that there are many more who could and should, but may escape the consequences,
As someone who has some IT knowledge, I used to build PCs for fun. I can tell you that trusting an IT system to work 100% correctly is ridiculous! Of course, it's going to have problems. A system of that scale built today wouldn't work 100%, let alone designed in the 1990s! As you say, its inability to record and report errors is a big issue. But so is the fact that the Post Office didn't seem to care that errors were not being corrected and, instead, blaming the Sub-Posts is the scandal. Jenkins' insistence that he didn't know what an 'Expert witness' actually was seems to be his only line of defence. So it's the 'classic' - Is he utterly incompetent (not knowing his court role) or a liar! (who knew his role and is now pretending that he didn't)
So sorry for replying so late. Excellent insight here. Yes, it's crazy we'd ever believe a computer would be 100% correct, given computers are designed by humans!
p
Thank you for these well described and articulated [also I think accurate] observations. NZ
I am so sorry for taking so long to acknowledge your comment. thanks so much for your encouragement. It's much appreciated.
Jenkins confidence in his own abilities is a classic programmer's spectrum response. His candid approach was as a result of being abandoned in the forest and assuming thinking he was walking on a path
So sorry for replying so late. Such a lovely analogy. I think there is certainly a kernel of truth in what you write.
p
Thank you for your insights
I am so sorry for my late acknowledgement of your post. Thank you for your lovely encouragment :)
Congrats on a thoughtful piece and promoting a thoughtful discussion.
your “move fast and break things” alludes to ‘agile’ software development which you reasonably well describe some aspects of but which would not not have been used for Horizon. Nevertheless Horizon like 99% of IT projects, according the evidence, was accepted into service by the customer and according to contract with faults remaining to be fixed and further faults remaining to be found. Your assertion that the product was “released to market” is mistaken as is the suggestion that there was no mechanism for finding faults or fixing them, it is highly unlikely that any fault would have been fixed without testing and acceptance by POL and released into service under the control of POL. IMHO POL’s assertions that they knew nothing about bugs errors and defects Is simply absurd.
I’m not sure that Fujitsu would have contracted for a system development whereby POL would prosecute its agents for failures in the system. I feel that Fujitsu were remiss in not controlling their clients actions (and that sounds stupid when you write it) but I cannot get past the actions of the legal profession in promoting and supporting this miscarriage, all the while professing to be simple agents of their client’s desires and simple mouthpieces of their client’s words.
I am so sorry for taking so long to acknowledge your comment. You make some great comments here John and you are right to correct me on those things. Thank you for pointing those areas out where my analysis didn't quite work. Great to hear your insights on this.
p
Very well put and I agree.
Thanks for your feedback, much appreciated.
p
My job was about 25% reducing the occurrence of harm to a a very small figure. When something went wrong, our first priority was identifying how to prevent it from happening again. But hey, what do I know, I'm a mere nurse and operating department practitioner.
I am so sorry for taking so long to acknowledge your comment. Great point. If only those folk worked to your standards. I think we seem to value those industries that churn out immoral actors and forget about valuing those sectors of society that only function because they are populated by moral actor. p
You make a lot of good points, but many of them only distantly related to specifics of Jenkins' testimony. I think you hit the nail on the head in your final sentences: FJ didn't design the system for the purpose of prosecuting postmasters, and when they discovered that their clients (POL) were using it for this purpose, they turned a blind eye. A lot of this is the culture of "the customer is always right", some of it underpinned by what's written in the contract: if the contract says you will support the client by appearing in court, then that's what you do. I don't think there was a "move fast and break things" culture here (you overlook that many of these events were 25 years ago, not 10 years ago), but there was a culture that involved lack of openness - sharing of knowledge with and between subpostmasters would have made everything operate so much more smoothly. The key causes of the problem in my view are (a) the adversarial supplier/customer relationship (which is normal in public sector procurement, and accounts for the fact that public sector IT projects are so expensive and unsatisfactory), and (b) the adversarial relationship between POL and the subpostmasters, which is a distinctive feature of this particular project.
Thanks so much for your posting Michael. It's really well thoughtful and I think you analysis is really useful. It very much adds to our understanding of things. p
I think that is a very kind, sympathetic and fair assessment. Mr Jenkins, I think, was strangely and, you might say choosily, incurious. He was poorly briefed by his legal people on the ‘expert witness’ role, yet seemingly failed ever to question his role, other than to say he thought he was on the prosecution side. He betrayed a certain vanity, for example, his pride over the ‘distinguished engineer’ badge. I think the reality is he was a lonely, long-time company man, and was easily flattered by the attention he got. He let it go to his head. But hey, enough of my yakking (Spinal Tap gag). Oh, last thing… please do a number on the government reps’ roles at UKGI and SHEX. Their culpability is striking, and will almost certainly go unpunished.
So sorry for replying so late Stephen. That is a great observation. Your word 'incurious' is perfect and might be applied to a number of the POL executive too. So, thanks for posting. If I can get to UKGI and SHEX I will. Is there a particular witness you want me to look at?
p
@@PaulDuckett No apology needed. Thank you for replying, and more pertinently, for your excellent posts here on this and other topics. I've yammered on elsewhere about the PO's own 'one of us' mentality; I'll have to revisit the inquiry catalogue to remind myself who's who. From memory, it's easier to recall who not to finger, if you'll pardon the expression. Anyway, thank you again for your great work.
I am a super geek and have been writing and releasing software for 30 years, there is no way horizon would have been released if I was involved. Do you watch / read any Dan Ariely?
Thanks for posting and I hadn't heard of Dan. I had a quick look and he seems to be a really interesting fellow. I will try to read some of his work when I can. Thanks so much for the suggestion.
p
Ahhhh... Dan Ariely and his questionable data.
@@humblepi3666 oh yeah didn't know about this. #Scandal #Scandal can't trust anyone these days. Back in the day you could leave your door wide open :)
What he did was trying to defend himself from the legal investigations that he is currently overcoming.
He can and will say that he was never properly instructed and having seen Jarnail Singh provide evidence he may well be believed.I also think POL changed his statements without checking with him perhaps because it was obvious he was really engaged with the Legal Process.POL wouldnt put him in the box ever again but did use his evidence which was rather naughty.
Yes, an absolute fair point and needed to be said.
Thank you :)
p
As you say his testimony was hard to assess
Overall, I would say he should have been nowhere near a courtroom, and did not have the skills to be a court expert
His work in this area sullied what was probably otherwise a very good career
Whether he was as he says 'trapped' into doing this work or not, he had agency, and he should have had the good sense to leave it to someone else better equipped to do it
So sorry for my late acknowledgement of your comment. Have just read it and I thought it was really good. It's great to hear your view on this and you make some excellent points.
Jenkins morphed from internal technical advisor for investigations, to internal Witness of Fact for prosecutions, to court Expert Witness but not independent crucially - incompetent guidance of legal codes to Gareth by POL internal lawyers, POL external lawyers, and judges. Lack of drilled down data analysis from Fujitsu, "tech snowing" by POL witnesses, and the unmerited respect of them by the courts, then roller coasted any defence who found it hard to contest or were ignored.
I came down on his side because two things. Firstly what he described in himself, I had actually seen in others when working. And secondly, he was not employed as in investigator of fraud and the like. So let me explain.
Firstly those I have seen acting like him, They had tunnel vision and were focused on the task they had been given, and were not able to really look and assess what was happening outside of that. Because they were so busy, that to get everything done that was demanded of them they just had to plough on.
And I saw this in him he was doing his day job, he was actually having to deal with problems that were arising as part of his real job then he was being asked to do this as well. Its more than obvious to me that nobody sat him down and told him exactly what the score was, there was no change or modification as to what he was doing at any time in relation to what he was producing. No attempt to change anything. You could clearly see where others were trying to make him say things and he was not prepared to say things. They said he was an unreliable witness because he had not disclosed things. But he had not been asked that question by anybody in the context of any of the trials or the situations. He had been asked to look at something in isolation by the post office. He was not specifically asked to look at the consequences of what he had found and if they had applied to other cases. Remember there were 900 prosecutions. He only provided evidence for 15 of them. Its the INVESTIGATORS at the PO that should have bene asking the questions that others that point the finger at him lay at his door
And I saw a lot of that In Gareth. He was being fed a narrative by PO staff who were focused on the brand, and that is where his comments came from about the postmasters. He had never met them he was repeating what was being discussed.
Secondly HE WAS NOT AN INVESTIGATOR. He was never asked to investigate the issues in the context of investigating a case of fraud. He was not asked to investigate what he had found could have caused in relation to criminal cases that he was not aware of.
The prosecution team, the po solicitors and barristers, the PO fraud investigators.. That whole process, that whole team were not being properly managed and coordinated properly at all.
I think that the system itself is robust but it does not have good checks and balances and the whole process of managing things with it was a sack of shitte.
I am so sorry for taking so long to acknowledge your comment. You make some excellent points here. Thank you so much for sharing your thoughts, you are very insightful and a nice bit of 'bugger it all' at the end that made me smile.
p
Great presentation. The issue in Software Development is compartmentalisation, and third party vendors being part of the life cycle. If you break down the process into individual sections / modules and pressure is on regarding timelines / cost , it is inevitable for systems are rolled out unfinished.
The only way to resolve these issues would be in-house monitoring teams that answer exclusively to the customer, and clauses in contracts enabling prosecution of developers .
Scientism is the new religion, and people believe it can do no wrong, and this leads to shortcuts, that impact on people
So sorry for replying so late. Great comments and great insight. I agree with what you say about scientism and you put it really well. p
I can assure you that whatever was going on at ICL, later Fujitsu, it was NOTHING like the Silicon Valley culture of 'move fast and break things'. This was a classic government/private IT contract and it failed exactly the way they all fail except in this case they managed to cover the failure up for a couple of decades. The British Government runs on fudge, or as they prefer to call it, 'creative ambiguity'. This is lethal when it intersects with computer systems which cannot tolerate ambiguity, and with competitive fixed price contracts which require a clearly understood scope. But the main culprit here is the legal system which somehow managed to convict people without the prosecution even demonstrating that any money was actually missing let alone that the accused had stolen it!
So sorry for my late reply John. Thanks for posting this. I think it's a good critique of my analysis and you do paint a picture of government that I think is better than the one I have. Nice work and thanks for posting this.
Maybe Gareth Jenkins was totally unaware of manipulative people and how they used him for their own ends. He was a degree carrying maths graduate from Cambridge steeped in code writing for computers. He did as he was asked by employers, too naive to see beyond the request. I think he got hung out to dry by everyone. A fall guy.
Hi Rodney. I am so sorry for my late acknowledgement of your comment. Great comment and you've summed things up brilliantly.
p
Thank you so much.
So sorry for replying so late. Glad you liked the video :)
We have a relative in the IT industry and she often says that the IT software guys seem so focussed on the system itself that they often cannot see any implication of anything outside their circle of expertise .
I am so sorry for taking so long to acknowledge your comment. Great to hear this insider perspective. Thank you so much :)
p
This is a really interesting perspective. I agree that there is a moral code with software across B2C interfaces being used to pass the burden of corporate administration back to the customers. When there is a problem it becomes a user problem not a code design problem. Agents are left in the middle and are given a script to obfuscate "you are the only one experiencing this problem". The Fujitsu system was released early and I have always suspected, predicated on Fujitsu quarterly sales targets. I also think Jenkins wrote the code and tested it, but there appears to be a complete absence of additional scenario tests he may not have been responsible for - e.g. lost comms link, hardware failure and security architecture.
so sorry for taking so long to say thank you for your comment. You make some very important points here.
Agree with your analysis! He can be seen as perhaps an accidental, reluctant, and malleable expert witness. The most knowledgeable about the code, but not about the customer. He saw himself on POLs side - part of the team, and therefore had a similar level of curiosity as POL...
I am so sorry for my late acknowledgement of your comment. You've put things very well indeed. Thanks so much for posting this.
p
Hi Paul, just FYI, Microsoft will freely admit, going back to the early 90's that they released software when they could sell it, not when it was fit for use and they have had some real duds too like Windows 8.0 and Windows Vista.
Back before windows server 2003, I earned a fortune in after hours call outs because Windows NT was so bad.
This is fab detail. Thank you. Windows have made some terrible moves. I remember windows 8 and how it tanked.
p
Thank you. I can normally make a judgment but with this I kept wavering. I think that he was naïve and believed what he was advised by information (or lack of) by Post Office, who were trying to cover it up.
Sorry for my late response. It was a tough one trying to figure where to sit when it came to Jenkins. I think your end judgement was a very fair one. Thanks for sharing :) p
I think your'e absolutely spot on , when you describe the users of Horizon, ie Subpostmasters, were more than likely and unkowingly being used as let's just say it ...'Guinea Pigs' ? as compares to your own analogy of Windows users etc back in the early 2000's to seek out all technical glitches ?.
Seems to me that both POL & Fujitsu are equally to blame for the carnage that subsequently resulted .
I am so sorry for taking so long to acknowledge your comment. It was good to hear from you and yes, I think you are right about spm being the guinea pigs.
The class structure sketch with John Cleese, and the two Ronnies cone to mind with your underdog psychology here! 👍🙏⚖️🙏
I am so sorry for my late acknowledgement of your comment. Loved you comment. It made me giggle :)
p
G" Day I have been following your comments on the Post Office and I like you originally a POM 1965 I came to Australia first Adelaide and now Brisbane I worked from the age of 43 until 73 with disability I had lots to do with your profession and found that they like Mr G Jenkins when trying to treat a person looked at the obvious behaviour and not the outer rim so to speak and had very closed attitudes so it's good for me to hear you without feeling that ha ha I am 80 yuck
Great to hear from you Lesley and thanks so much for posting. Good to know there's another POM nearby (though I am up in North Queensland!) Yes, that is one thing I do criticise psychology about too. They look at the obvious but then they put it through a meat grinder so it comes out as indecipherable mince meat and then they try to sell it back to us as scientific understanding. Well, that my warped view of things :)
p
I have watched quite a few of the inquiry videos now and what sticks out to me is that everyone was at best lazy and incompetent and at worst ruthless and arrogant. echos of the 2008 financial crisis? of course the software didn't work as it should cos every piece of software is like that so to stand up and defend it for decades without properly investigating beggars belief (and I not convinced they could investigate properly either cos they didn't seem to have what should be a basic and fundamental mechanism to capture all the errors or report on them). I think people also have a natural tendency to rally around and defend blindly when things go wrong ! and you can still see that today in the defensive sitting positions of the Fujitsu and PO execs - and how strange is it they can't remember things if its about themselves but they can remember quite clearly if it related to someone else ! also Fujitsu would have been worried about not being paid for horizon work if there were problems so they weren't unbiased here and would have proactively buried their heads in the sand - its like asking someone to mark their own homework! from the post office side they were spending millions on these trials and no one thought to ask questions if this was the right approach - they would have been better off putting that effort and money into listening to the problems and fixing them - this speaks to a toxic ruthless culture. i have to say as well I am not sure about the value of the inquiry it just seems like a bit of theatre to appease the masses - cos nothing of consequence will happen
I am so sorry for taking so long to acknowledge your comment. thanks so much for posting. Yep, I do worry that this is just a piece of theatre with no action at the end. I really loved your analysis of the situation. Thanks so much for sharing.
p
I doubt that the Horizon computer system is morally bad - its just a fancy networked shop till system
The key decision was how to manage fault rectifications
The system was buggy, and required huge expenditure to support it in-service
When problems arose, PO needed to rectify them and support postmasters
Instead they chose to blame postmasters, and prosecute them
This was disastrous because the system could therefore not be rectified, because noone could admit the truth, and would inevitably blow up into this huge enquiry
Jenkins unwisely got carried into it I think, rather than openly fixing software issues, as he should have been
To me thats the moral bit of this
I am so sorry for my late acknowledgement of your comment. You make some excellent point here. Thank you so much for posting.
p
Some time ago there was an email (I think from PO general council) saying that Gareth had not been advised of duties as expert witness. Also that when he gave evidence at Misra he was not considered an expert witness (so wouldn’t have considered that note about it). There was a problem that the judge asked the expert witnesses to have a meeting and he was included. They then realised he had become an expert witness accidentally. I’m surprised this wasn’t pursued in this phase.
The PO negotiated the price down so some of the visibility of tracing was dropped and information available from Fujitsu was restricted without extra payment.
I think things would have been traceable (by Gareth) but the PO didn’t have an incentive to pay for that, Fujitsu didn’t want to do it for free and the PO concealed from the subpostmasters that it could be available.
I think he was probably being honest - he was just interested in the system and was limited in what he was allowed to investigate.
He should have been more obstinate about what he was willing to (not) do but had been with the company a long time and was probably concerned about leaving.
I’ve been in a similar position and think it odd that he is meant to be in a senior (expert) position but doesn’t seem to have input into what work he does. I usually tell companies what needs to be done rather than the other way round - how can non-technical managers know?
My opinion anyway.
He’s just interested in the system and it was some else’s responsibility to use the information he provided.
PO was just interested in info to support their position, Fujitsu were just interested in profit and keeping the PO happy cheaply.
Good summary. Last paragraph great summary re GJ.
I am so sorry for taking so long to acknowledge your comment. Really loved reading your thoughts on this and you make a lot of sense. Thank you so much for posting.
p
Good talk. I do work with technology and can tell you have hit on a lot of truths about the changes in culture of testing. Regarding Gareth Jenkins, one of my roles is often dealing with the ego of developers and architects. These people are often detached sociopathic individuals. Have very one dimensional lives, they block out the reality they can make a mistake. Trouble with people at genius level they are incredibly blind to the idea they could make any mistake.
I don’t think he’s genius level. I found him pretty unimpressive.
@@martinbaker6532 some people think they are, but yep thinking you are and being are two different things.
So sorry for my very late reply. I am still catching up on the comments. Great post, you make a great point here.
Just a suggestion. I had a sense that he was on the autism spectrum . . .high functioning but not given to emotional expression?
I thought that too.
Yep.
I agree, the question is, would this give him a “get out of jail free” card?
@blyndaza good question. I think not. My comment was primarily aimed at the comments by Dr. Paul about Gareth not seeming to care about the submasters' plight. I offered it as one possibility to explain Gareth Jenkins' apparent lack of emotion.
Hi Arthur, yes, I think a number of people would agree with you here. My take on this is that we are all different, some of our differences are exagerrated by the circumstances we find ourselves in and perhaps this is what is happening here. He did seem like a fish out of water.
p
Placing moral responsibility, or lack of it in this case, onto the computer system is in in my opinion intentional. It’s a perfect way for management to step away from their own responsibilities. If anything goes wrong, if anyone is harmed then they can say it wasn’t me, it was the computer.
In the case of our friend Gareth, I started his testimony thinking he was the PO’s useful idiot. But, as the days went by, I could see more intent. His lack of moral compass was being exposed. He was on message with PO lawyers. Still the usefully idiot but completely on board with trampling the little guy into the proverbial dirt.
So sorry for my very late reply. I am still catching up on the comments. Really interesting take on this even though I got squimish at the thought of trampling on the little guy.
p
A great well-framed analysis Paul, well done. With the title of Distinguished Engineer, how could Mr Jenkins behave without considering moral consequences of the situation? Maybe he also failed to understand his obligations to the title? Moreover, I would posit that he failed to either abide by, or understand, his own institution's code of conduct? Just a thought.
Thanks so much. So sorry for my late reply. I think that is not just a thought, I think it was an excellent thought. Thank you so much for posting.
p
In Mr Jenkins 4 day witness to the inquiry there were many instances and regular repetitions indicating neurological challenges. If this is the case the evidence may need to be filtered through an understanding of how this would affect all interactions, subordinations, articulations, vocal expressions, audible reactions etc. People all have limitations on how we gauge personalities due to our own experiences and analysis of others. Therefore, not all people may have the ability to recognise the signifiers or understand how it could affect and condition conduct with and around others and may lead to a subsequent misreading or misinterpretations of evidence given.
So sorry for replying so late. Such an interesting set of thoughts here. Thank you so much for posting, you make some excellent points.
p
Technology has dehumanised ethical practice, governance & the simple trait of honesty. A rapid erosion of 'proper counsel & sound judgement' is a rarity in big corporations. Decisions are too often based on someone else's opinion of events instead of hard facts. Investigative processes only kick in when the damage has been well & truly done. It's all the wrong way round - our unhealthy reliance on the 'logic' of technology has seriously diluted good conscience.
Love the brain-workout episodes you present. Keep going.
I am so sorry for taking so long to acknowledge your comment. You make some excellent points and glad that my videos give your brain a workout :) Thanks for the encouragement. It's much appreciated.
p
As a former contract manager I have always hated the adversarial form of contracts, rather than a collaborative style. I know it’s very hard to achieve. It does not serve either party when severe and threatening clauses make it impossible to be honest with the buyer about issues. This clearly happened here, the PO screwed the price down so much that the operational side couldn’t afford the testing and monitoring involved. They couldn’t speak openly about issues and it encouraged them to lie about how the system was working to avoid penalties etc.
Like so many government I.T. projects.
Love this post... glad to see you posting it twice :)
p
It would be interesting to get your take on the illusionary truths operating during the scandal and how various people in the different organisations involved responded to cognitive dissonance caused by evidence that challenged them. The obvious example is the assumed venality of the SPMs, but there were others, including the idea that Horizon was always operationally sound at a local level because it operated properly at scale. It feels like uncomfortable facts were repeatedly shunted into a “zone of irrelevance” through whatever mechanism was most naturally suited to the person involved, whether they were managers, technical experts, or part of the legal system. The degree of conscious agency that each had when doing this is something that the Inquiry seems very interested in.
Martin, I am so sorry for taking so long to acknowledge your comment. This would be such an interesting angle to take. I love the way you have put this - particularly love your term' zone of irrelevance'. If I can bring this into a future video, I most certainly will.
p
@@PaulDuckett Thanks Paul. It would be great if you do. That zone seems to be worryingly easy for our legal system to access!
POL were a knowing party to someone who passed himself off as an Expert Witness, when POL knew full well that, even after being issued relevant legal warnings, he persisted in passing himself off as an Expert Witness
I am so sorry for taking so long to acknowledge your comment. Loved you post, you make a nice summary of the issue. Thank you.
Good analysis and interesting viewpoints. One aspect that you don't mention is that not all software is created or used equally. If Facebook fails and a comment on a post is lost, that is not a big deal. When Horizon failed, someone took their own life or was sent to prison. As for Jenkins, his explanation that he didn't understand his role as an expert witness is also BS, or does that apply to other parts of the law also, i.e. if you "dont't know" that killing someone us illegal, then murder is OK? As an IT system guy since more than 40 years, that is appaling, my morals are WAY higher than that. His excuses are 100% non-acceptable and he is a disgrace of an IT-pro.
So sorry for my late reply. You point out a really important aspect. Thanks so much for posting this. :)
Paul - software for 'safety critical systems' is subject to much more engineering rigour (i.e. standards, level and scope of testing, audits etc) as compared to the Horizon accounting software, as these can potentially result in the (more obvious) loss of human life. Accounting software isn't generally considered as 'safety critical' - however the actors involved in this case clearly brought about a loss of human life of another sort.
Really good point. Thanks so much for this. You make a really important point.
p
I'm a 1980s Computing Science graduate who's worked on many software projects, and I couldn't agree with you more about the 'morality' of IT development. Coders rarely realise that they're coding in their own assumptions, and these assumptions often come from a limited model of the world. Error-handling is critical; but developers frequently lack the imagination for all the things that could go wrong. My adult son is convinced that AI will be wonderful, and I argue endlessly that it's essentially just code, and that the code is written by (mainly) young men with narrow life experience, and that the built-in biases will colour the whole field. I've watched it happen again and again. Gareth Jenkins seemed to me just such a narrow-visioned technical wonk, absolutely sure of his rightness, and therefore very useful to the legal sharks who latched onto him.
Thanks so much for you post and sharing your experience. Really valuable to here this. I agree with your assessment of how Jenkins may have been used by the legal sharks.
p