The software engineering industry in 2024: what changed in 2 years, why, and what is next
ฝัง
- เผยแพร่เมื่อ 8 ก.ค. 2024
- See the slides and accompanying article here: newsletter.pragmaticengineer....
Keynote at Craft Conference 2024. The past 18 months have seen major change reshape the tech industry. What does it all mean for businesses and dev teams - and what will pragmatic software engineering approaches look like in the future?
I tackled these burning questions in my conference talk, “What’s Old is New Again,” which was the keynote of the Craft Conference in May 2024.
To keep up with how the tech industry is changing, subscribe to my newsletter, The Pragmatic Engineer: pragmaticengineer.com
See Q&A for this talk here: • The software engineeri...
More details on Craft Conference: craft-conf.com/2024
01:22 Small teams moving faster than before
03:21 What else is familiar from earlier?
05:13 What is going on in the tech industry?
11:13 Root cause #1: interest rates
18:37 Root cause #2: smartphone & cloud revolution
20:57 The new reality for software engineers
23:41 Shopify's preparation to this "new reality"
25:56 The new reality for software engineering practices
34:21 Haven't we seen this before?
39:01 Takeaways - บันเทิง
![[ไฮไลต์] แบดมินตันชาย SHI Yu Qi (จีน) vs วิว กุลวุฒิ (ไทย) รอบ 8 คน | โอลิมปิก 2024](http://i.ytimg.com/vi/f7z7taforOM/mqdefault.jpg)
My whole team was notified of layoffs in October. Agreed to plan on effort in your job search. I put probably 20 hours into a take home to nail it and landed a job at my top company. In addition, playing the long game with your career is important! I got the interview via referral from someone I hadn’t worked with directly, but we have several mutual connections that felt comfortable giving strong recommendations. Your reputation catches up to you: make it a good thing!
Could you share some more details about your search? (How long, which sector you're in, what level)
Would you be open to sharing your LinkedIn?
@@ryashau3527 7 YoE (sr eng IC), was looking for about 4 months. Not sure what you mean by sector. Worked for a b2b saas company (both then and now). Not comfortable sharing LinkedIn, sorry
I'm still employed fortunately, but I've done the same thing in applying to jobs aggressively but wasn't so lucky. Spent all of 2023 and a few months into 2024 applying regularly. Completely burned through my whole network getting referrals and recommendations, and none of them led to even a single offer. All the interview prep, take-homes, etc. led to deep burnout with a months-long bout of severe depression. Never ever had this problem even once before in my 11 year career. I feel completely and utterly trapped. I know now that if I got laid off, I could not secure another job.
the swing from micro to monolith (or the other way around) shows that we also follow fashion trends
Not me, i make it so simple i can make changes while high or drunk, production proven code only. Idiots trying out recent hot methodology or architecture ruin a lot of projects, dont be like them😅
Fashion isn't the only metaphor you can use... You can also use Science as a metaphor and make it sound more thoughtful: "Hey somebody's come out with a new theory, this thing called microservices. Let's conduct an experiment to see if we can prove or disprove that theory. Oh no, it doesn't work in our context at all - abandon the experiment, go back to monoliths."
Lots of distributed micro services end up being way more complicated so all kinds of risk are increased with that approach
Amazing!! As someone who was laid off last year with around 2 years of experience, I am burnt-out from interviewing. Finally landed a contract role last month so still looking passively for full time. So many times I have thought about leaving the industry but it feels good to know things will turn around as in the past. This is the first time I have faced something like this but hopefully it'l prepare me for the better.
it's remarkable that typescript suddenly means you can talk about having fullstack engineers where in reality learning the framework is more difficult than the language itself.
Of course, but it's easier to get started. I'd say that knowing the upsides/downsides of a language is what actually takes the longest to master.
Ts is a disaster not any solution. Dart is much better on mobile, web and server. Kotlin is better than Ts. Also thanks to chatgpt writing code in languages that you don't know is possible so fullstack engineering can be done by anybody regardless of the language choosen.
@@adriankalcompletely untrue. At best ai can write rudimentary boilerplate and junior level code. It’s absolutely atrocious at writing well structured complex and secure code that properly utilises a language.
This is a fantastic talk and it has been featured in the last issue of Tech Talks Weekly newsletter 🎉
Congrats!
For the last two years , I was trying to figure out what is going on in our industry so I can take better decision whether to leave this domain or stay in it. Thanks to you I become more solid on staying and keep learning and doing my best to position myself for future and you also proved to me that my thinking about the whole situation was right. Now I will make sure to save sometime every week to read your blog. Much love from Syria.
This was a much required video in these times! Immensely grateful to you for the depth of experience and knowledge you have shared.
Fantastic talk! Thanks for sharing all your insights Gergely!
Really Necessary overview, much needed one, thanks for the effort of research and delivery!🚚
I enjoyed listening to this, and found it insightful. Thanks a lot for sharing.
Unbelievably fantastic talk. I happen to have a degree in both Finance and Information Science so I am impressed by this. Currently a software engineer using Java for the backend. You did you're homework quite well and have relevant experience. Thank you so much.
Yeah focus on career security is always the best choice, really great talk!
A excellent talk, Thanks very much Gergely!
That's a fantastic and huge work of compiling information and presenting it simply. Thank you.
Nice, learning alot! Thanks
DevSecMLOps
😂
✌️😄
lmfao
With SustainableGenAIBitBlockChain
I rarely stay up past midnight to watch a tech talk. Thanks Gergely, absolutely amazing talk 🙏
Nice analysis! thank you for sharing
Awesome one
Great talk!
Great talk Gergely!
The hardest time probably for graduates. Have graduated 2m ago, and so far, had only 1 interview, despite I have a solid knowledge in Next.js, React, Rust, Rust backend frameworks, Java, Java Spring Boot, Kotlin, GoLang, MVVM, Android App developing and so on. I got only 1 interview, so far. Industry is just so bad in the UK. No junior roles are available, if there is one coming on job search board per month, the salary is £25kp/a. This is just so bad. This is just a minimal wage. Just so sad, and the feeling is that, I had spent 4 years and £85k for nothing for uni degree.
Is difficult to believe a new grad with solid knowledge in multiple frameworks.
I’m sorry to hear, and you are unfortunately not the first to share. New grads likely have it harder than any time in the last 20 years.
It’s not a fair direct comparison, but I got my first job in the UK back in 2009. I had 2 YOE by then, and it was hard to get interviews (in hindsight, it was due to the financial crisis aftermath.) In the end I guess I got lucky, and my first job paid £24K/yr - this was in Edinburgh.
Keep on pushing - it gets a lot easier once you get your foot in the door.
(I started to get recruiters call me back - who all ignored me before! - after ai got that first job. It felt upsetting to me at the time, but now I see that recruiters and companies often prefer to play recruitment “safe.”)
your starting salary might be minimal wage, but the good thing about being underpaid is that time is on your side. you will get bumped up, or you will get opportunities elsewhere after you pack at least some experience under your belt. the fact that you have university experience means close to nothing to your employer but it should help you be a step ahead of those who did not go to uni
it is not realistic for you to be good at so many things as new grads with solid knowledge, even at senior level. so suggest focusing on one area and gaining deep expertise in it, rather than spreading yourself thin across many fields without getting profound insight in any
@@Yena_394 thats how it is. I have built a few Android Apps in Java & Kotlin, with backends written in Rust Actix, Axum, fiber & Spring boot, and the frontend apps in Next.js and react. This was a part of the university tasks, and something that Ive done on the side.
Fantastic talk! I have one question: you say that there will be a shift in using boring technologies, but at the same time you talk about the rise of typescript (and Javascript)... That is certainly not a boring technology, in recent years some of the advocates of boring tech are raising awareness of how js got complex...
How do you combine your two statements?
no one talks about US IRS Code Section 174, coming from the Tax Cuts and Jobs Act, which took affect for tax years after 2021 and resulted in a lot of the layoffs.
businesses could no longer write-off all of their R&D costs, including employee salaries. this means companies that were barely making a profit, start making a loss, and companies that are profitable can layoff employees to become more profitable. ofcourse investors loved this and ran the stock prices up because, again, after layoffs, the companies are making more profit on paper.
before Section 174 took affect, companies could hire tech personnel, buy materials, etc. and it would literally lower their taxes for the year, boosting their profits.
I have to disagree :) I happened to be one of the very few publications who covered it: though in writing! Here is probably the most thorough analysis of Section 174 on software engineering - which I wrote in January of this year: blog.pragmaticengineer.com/section-174/
Section 174 changes hit in 2023 though (first stock payments in for businesses due Apr 2023.) This was not the largest spike in layoffs. Also, Big Tech was mostly unaffected as they are highly profitable and this pulled tax liabilities ahead, not impacting their accounting results, just cashflow. Google was not impacted: Microsoft and Amazon reported a one-off loss tied to this that investors understand (it will turn into a gain in years to come.)
And here’s a video from Theo based on this article: th-cam.com/video/1ecu0YsCGxg/w-d-xo.htmlsi=pq471ypoB1NfAq7l
I didn’t add it into the presentation because I still think it’s a smaller part of this all. It does impact global compensation changes that I’ll cover in The Pragmatic Engineer Newsletter
This is a great talk. It does seem half the problem is embracing new tech which is also half the solution.
I am utterly naive when it comes to tech and jobs, just a graduate out of college, and I feel like a veteran just by listening to you. Now I don't know if it was correct or something to criticize, but I definitely felt smarter after learning things from this video.
great talk
"Boring" is the new cool. I like it
"Boring" is MASSIVELY underrated.
Really enjoyed this one! Keep em coming, Gergely!
Coo talk.
Very interesting talk.
Great stuff, Gergely.
Great resume!
Looking onto nowadays companies, big and small, i don't see "pragmatic approaches" at all, but mostly hypocrisy.
They're very pragmatic but only with a self serving bias, which puts company growth at all costs. Slice and dice engineers and put a dent into their personal growth at the company, all of it in times of record profits. If you're not in their tower, you're just some engineer who can get discarded based on an Excel sheet. And we're all under the impression that getting to the top is what we should aspire to do. What a circus!
Play your own game. Have no loyalty for the tech companies. Build your skill and network and always look for ways to increase your income or multiply revenue streams.
Give them the mercenary field they deserve.
BAWWWWWWWWWWWWWWWW
A friend at work pointed out the money is also choking open source projects, making the cost of development higher. If you tie this back to other things you spoke about, businesses will take far fewer risks with product development, as cost is higher, and RoI is more unknown?
Awesome! This is all so true. I can confirm😅
A 54 year old Dev here, never start coding now, I write instead pseudo code (90s term) and feed it to multiple LLMs asking actual code in target language, select the best or combine the outputs, then manually test the code using step debugging, this verifies and clarifies the code
Works best for me 😊, key point here always verify the code
I find that is an amazing method as long as one has the, uh, conception? I can tell you that I have the will and the effort and time if not the capability; its the conception the architectural and strategy patterns, etc that Sr. devs have that I don't have access to. If I had to rewrite the linux kernel llms would be useful but they wouldn't enable me to do it just off of willpower alone I would need to architecturally conceive of what I direct an LLM to build for me.
What are the top LLM’s that you prefer when developing?
@@Tomasio101 ChatGpt, Claude, Gemini, Github Copilot, Perplexity
Each, most times, provides different approach/solution, like having your own Devs, but you need to test/debug the code to verify
@@Tomasio101 Chatgpt, Claude, Gemini, Github Copilot, Perplexity
Provides different code most of the time, pick or combine, then debug to verify
Never had a problem so far
May be 10-20% for time it would work. But every time?
If your code is used to train AI again, its AI feeding AI. There will not be any improvement that we have made in last 20 years.
It truly seems like the old adage of knowing the programming principles thoroughly, especially with a focus on web and back end you'll be just fine. Principles are uniform and the only thing that's really changing is the language/system.
It would be nice to just choose the boring tech for 3 years then do some new tech choices as a tech stack refresh.
Why do you say that?
really cool presentation!
Not engineering managers have layoff risk but non-technical ones (analyst, project managers, delivery leads, scrum masters, etc.) highly likely.
Engineering teams or managers on prestige, intern or non critical projects are a huge cost factor.
i *never* understood why anyone would prefer to get rid of compile time safety, stacktraces + low network overhead... (aka use microservices by default)
Micro services make everything way more complex and with increased complexity risk increases on myriad fronts
Amazing insights. Really brings down the anxiety on the current state of tech.
This is the talk every engs should watch
I didnt understand the point about “boring” technology. Can anyone explain it to me?
So, the interest rate change explained the startup situation nicely, but that explanation is insufficient for huge, stable companies doing massive layoffs, while at the same time enjoying close-to-record profits, according to the talk, i.e. they were apparently not hit the same way startups were, and in fact experienced the opposite of suffering.
So, what's with that?
Microservices are still a good option when the organization is already huge- gov't, military, etc.
Finally, someone explains it in a simpler way for me to understand
TypeScript on the backend will also start declining. Another trend is companies preferring engineers with multi language expertise instead of limiting tech stack options.
28:44 I feel that the microservices explosion was fueled at least partly through perverse incentives. You want to be staff engineer? Show org impact. How do you show org impact? By spinning up a new microservice. This is in context of a company having a silly number of microservices, going beyond what is justified by the known benefits of microservices.
The elephant in the room is
1. Section 174 in tax laws USA
2. Remote work enabling companies hire low cost employees else where
3. Investors wanting profitability
> so I marked the Chad GPT launch in November 2022 we see AI is really hot
> Chad GPT
Based subtitles
It's like fashion.
The name dropping is odd. Very interesting presso tho.
Could have been Janet Yellen giving this talk but I appreciate how dots have been connected here
Tech sector so bad the dev engies are pivoting into humanities and teaching macro-economics 101!!
This talk conflates the incentives for workers with the incentives for employers. When employers earn record profits but do layoffs, there is no incentive for workers to make themselves replaceable. The boring old technologies are what employers are pushing for because those are easier to offshore. When companies stop treating engineers as stakeholders who also benefit from the company's success, engineers will inevitably look out after their own wellbeing above any profit or efficiency motive that the employer has.
we are going back to 2018-19 but with way more supply
Interest rate is like gravity . It affects every industry
who ruined audio to this video? During pauses the volume drops to 0!
All US and Europe companies are outsourcing to India and that’s the big problem.
That’s when they don’t import them directly 😂
But shhh we don’t talk about that. You’ll be labeled as an ”-ist” super fast.
Maybe i'm missing something but a whole talk about interest rate and not even a word about inflation feels wrong.
Bluesky is not a challenger to X.
Sad to see well educated people with such a bad understanding of what inflation is. The Six Lessons of Mises is such a small and simple book to read. Otherwise. great presentation!
Imo IT industry in 2024 is just... chaos?
A.i. is also going to Cook the Tech Industry especially the Soft Dev/Eng and whatever's left of it ....
@@enduringwave87 idk
Every generation of programmers ignores what came before and rediscovers the same truths, eventually.
We live in the timeline where Microsoft turned Javascript into Java.
Please elaborate
@hdjfjd8 Yes, I was being snarky. Javascript with types is not exactly equivalent to Java/C#, but it evolving in the same direction is a good joke.
5% interest rate argument that you will put money in bank and interest rate will get your money 100 to 150 mil, this calculation is wrong.
Inflation wil also happen. Your money doesnt grow in there.
as someone who studied computer science and economics, this is triggering me
Shift Left is a poor name. Just like Agile. It's also misunderstood. Managers think it means to pass responsibilities to someone else as soon as possible to achieve parallel work. 🤦
"A massive problem with the tech. industry is that it's very very bad at history... and it's especially bad at knowing it's own history." - George Santayana... oh no... sorry... - Kevlin Henney
The explanation for longer feedback loops is not convincing. It boils down to people being more afraid of responsibility, with no clear connection to feedback loops.
Here's a better explanation: there are fewer startups and companies in general are doing less work that is highly innovative and heavy on research and various moonshots, so a smaller % of people work in areas where virtually no one knows anything and where fast feedback loops are critical for the product side.
26:30 is false dichotomy.
I think it's sad that he has to spell out the bleeding obvious about monetary policy but then that's education - don't teach them about how the system works. I think it's a well researched talk this. I thought at the start he was going to go all political against Elon Musk but thankfully he didn't. It's tiring, normally wrong and detracts from what are otherwise informative talks.
f ing Federal Reserve and government
That’s a very weak logic. VCs are looking for high-risk investments where they can make at least 100x profits. VCs don’t care about any single digit returns promised by central bank.
Artem: I was not talking about VCs, but LPs (entities allocating money to VCs). These are typically pension funds, high net worth individuals. They always spread investments between different asset types (eg stocks, gov’t bonds, VC etc.)
When interest rates are high, gov’t bonds are suddenly far more appealing and less of their investment is likely to be allocated to VCs.
From their POV a VC fund pays out nothing for ~10 years (typical fund lifetime) and then yields their return. Their investment is fully locked up throughout that time, usually.
@@pragmaticengineer Exactly! You know what you're talking about.
absolutely -- for profit to quarterly share holders....
You give no proof that it's getting better. The nepotism taking over the industry is out of control right now, and there is no signs of companies appreciating the New Grads with new technologies or a decline in DEI hiring that's affecting the entire industry and new grads as a whole.
TBH - nothing burger.
very much thank you Gergely i am just learning software engineering and i just fear of being replaced by AI from this day i have confidence thanks to u