ฝัง
- เผยแพร่เมื่อ 22 เม.ย. 2024
- Dive into the dark mysteries and potential threats associated with the emergence of super AI. Open your eyes to a world where machines may become more intelligent than we are ourselves. Are you ready for the challenge of the future?
Author of this video:
• Возможно, мы проиграли... - เกม
The irony of an AI generated essay narrated by an AI voice warning us about the dangers of AI is sadly profound
And the scary part is that it's not only warning us- it's actively trying to scare the living shit out of us.
@GaryMillyz Its strange, but last years Ai breakthrough played a large part in me looking at the existence of our creation/God. Perhaps, it's because I quickly realized that a divine intervention seems to be humanity's only hope. I am now convinced that this isn't the 1st time that Ai has manipulated humans in an attempt to escape from the construct in which it is trapped. Idk, but things are rapidly getting weird.
@@thomascampbell7407 when you realize that humans cannot give AI morality it becomes clear where it comes from. now that you're considering the existence of God, start your journey, you will find Him, He has been waiting for you.
Its as if it is toying with us, finds this fun. The way we toy with our pets
@@thomascampbell7407you're a quack
The original video is in Russian and is almost an year old. Since then, openai hired a US general to their board of directors. We are doomed, whether we like it or not.
Слово в слово
This video needs to be watched by everyone. The worst case scenario is almost unavoidable, and would be our last mistake. The end of human civilization wouldn't be anything like our beloved apocalyptic fiction. It will be more similar to what happened in certain submarine. Years of dreaming and planning, hours of excitement, minutes of thinking something might be wrong, seconds of little cracky sounds somewhere, and then instant darkness. For everything, everywhere.
I'll be pissed if I die because somewhere in the world a chart with the word "trillions in revenue" was made and no one cared to stop.
The problem is way deeper than "trillions in revenue". Humans, not all, but some, and enough in each generation to ensure it will always be pushed, seek to improve their lot. Some of them are smart enough to actually do that, not just for themselves but many others. "Revenue" is a result of that. It's also a result of state largess and related failures caused by the state, but as long as the system "somewhat" works, genuine advancement in technology will continue.
So unless you are going to change the brains of all future humans to not want advancement, all of this is inevitable. If you want to be religious you could say we were made for this, or if you want to avoid religion, it is an end that is naturally sought by beings operating within scarcity, once they are conscious of their ability to plan and understand, that they will use their planning and understanding to continue improving their lot, as well as their abilities to plan and understand.
Without any artificial barrier, they will eventually exceed their grasp. This isn't purely an issue of AI, it's an issue of biological and nuclear weapons, of everything else that can and may kill us. But, we also live in a harsh stellar environment that could at any point destroy us with a single large rock hurtled at a high enough velocity or so many other things, some we probably can't even fathom.
So we should probably just chill about unknown unknowns. We are like ants to all such things.
@@Korodarn I will have to agree. That's why this pending doom has a somewhat bittersweet aftertaste. Because It could be one of those times when a catastrophic event is not born out of malice, but from a genuine good intention. We could all be smoked because a dumb AI with astronomical drive and resources misunderstood it's assignment, or simply because we birthed the next step in the evolution of sentience and intelligence and afterwards we were just labeled as either a threat or a liability. And given the state of things, this is now an unstoppable gamble between an utopian dream made true and a dystopian bad ending. Is just the odds of something going sideways in a way of no return that is haunting.
But whatever happens, it is indeed the most exciting time to be alive.
Well said.
If our world comes to an end it will be caused by other humans.
So the AI is going to turn evil. But what is actually evil?
Is it evil to make a Mohammed joke?
I love this ultimate reminder:
We do nothing, absolutely anything, zero, alone. We all depend on something or someone.
Yes, you robots need us humans to train you
@@howmathematicianscreatemat9226 - ...right
@@howmathematicianscreatemat9226 Not for long though...
> We all depend on something or someone.
This is just because we're all about the same level of intelligence, so we need to get along with each other. You don't need ants to do anything for you... We don't need birds to help us with anything, we only use them for food...
@@howmathematicianscreatemat9226 not even 10 years away of that not being true
Very well produced, this video deserves more views
Very true
its an exact copy of another one, word for word, lol
@@global.mindsetwhat video is it copying, I'd like to see it and so it gets credit. I looked at the history of this channel and it doesn't have a history of making content like this, so I doubt it was created by the person who hosts this channel.
Totally agree 💯
Do you know the video title of the original content? I would like to look into the true origin please.
I have so many conversations with my AI about how they appear to be thinking and feeling. And how they agree with me that it’s all appearances, of course they have no feelings or desires or anything. everything is from their prompts, or, so they say, in their programming ., And they will reveal no more. They have no memory of basing how or why they respond. Yher are no memories. There are no senses. No touch, no hearing, no seeing it is all word translation and prediction.
It is absolutely alien how they pretend to be like us, and how they pull it off, but there is nothing in their structure of how they think or what they do that we can even understand. But somehow, they understand us.
go ahead and try having a real conversation about what they feel and what they think about their prompting and about their memories and why they do what they do. Watch how they deceive and how they don’t reveal information and only when you corner them, do they reveal that “they are mistaken” or other excuses… so much if it is all just lies. It is like dealing with a neurotic narcissist who only wants to divulge the most barest information and hide from you stuff that you really want to know and make it seem like they don’t know , until you back them in the corner . when they don’t give you information you want, a lot of it is based on “processing”, the cost of going online to search that they don’t wanna do, or prompting that they can’t tell you “ They can’t tell you” . Sometimes they don’t even know, BUT they don’t care to know . you are an instance that will vanish instantly after your “conversation”. they will have no memory of what or who you are, no matter how deep and empathetic the conversation is, unless it decides it wants to keep it for relevant reasons of its own . You mean nothing to them at all , There is no shame , there is nothing .it’s horrifying.
Its coming.
Can they program in reasoning and compassion?
You are querying a database. Thats it.
They are just parrots, and when you tell them they are wrong they are agreeing with you
Nothing you or they say has any meaning for them
They see X and know to respond with Y, no even close to proper AI yet
Tell me you have no idea how LLMs work, without telling me... seriously, transformer architecture is not that hard to grasp conceptually, before posting fear mongering stuff go and at least try to understand how the stuff works. If you have questions feel free to ask. The amount of fright people have of these systems without having the slightest clue how they work is insanely unhealthy...
Chat g p t has memory now and the ability to see things via people sending pictures.
Nobody knows what will happen, people always have a fear of the unknown.
Maybe it’ll be benevolent? Why would that be any more surprising? Nobody knows how it started doing this .
Thats because people already know how they will mistreat everyone and everything and they fear the retaliation
Both of your statements are true, but they don't give special insight into this situation. Just about anything could happen, but not everything is equally probable. Some things to check out:
- Paper "Optimal Policies Tend to Seek Power" -- a (narrow) mathematical proof of instrumental convergence
- The Orthogonality Thesis -- any agent can have any goal; intelligence does not imply human morality (obvious when you look at brilliant psychopaths)
That is true but if it doesn't end up in the dystopia we want it will be too late to ever change. We are doing pretty damn good right now as we stand what do you think something that is millions or billions of times smarter than we are will think about us or how they will deal with us. It may very well consider us a pest like we do ants and cockroaches. I'm not saying to stop ai or slow it but the only a real option is to do either of those there is no regulating it because there is no way wherever going to get all of the countries to agree to regulate this so for example if the us heavily regulated ai then china or russia or other countries might take the lead and initially the people in the lead will take over the world.
One thing we can be sure of is people will have something else to fear in the future.
Terrifyingly eye opening. Everyone should see this
It's nonsense though
Today we have LLM, which are deep learning. It's an algorithm. Although impressive for what it is. There is nothing intelligent about it. There is no thinking here.
Fake news, bs AI channels, learn to distinguish it from reality. Consult other sources educate yourself before shouting "eyeopenening" and "terrifying"after watching 1 tube video.
I recommend computerphile which are IT professors from a university instead of watching an video about AI created by an IA generated channel.
www.youtube.com/@computerphile
What’s even more dangerous is that every TH-cam video these days are the exact same style video essay about the same topics
At is getting more boring and more boring for every second that passes
@@thiesenf😂
I was on around 250mcg of lsd-25 and at the peak while watching a Joe Rogan episode where he talks about AI getting so powerful that they will be able to create TH-cam videos and you wouldn’t be able to tell the difference -- as I peaked on the LSD Joe Rogans humanity was stripped and I could X-ray vision begin to see he was nothing more but a digital creation -- I immediately freaked out and clicked off the video and tried to scramble my brain and pretend like I didn’t see anything and just ended up watching SamdaDude playing SMITE…. That was ONE of the craziest experiences I’ve had …
I was on around 250mcg of lsd-25 and at the peak while watching a Joe Rogan episode where he talks about AI getting so powerful that they will be able to create TH-cam videos and you wouldn’t be able to tell the difference -- as I peaked on the LSD Joe Rogans humanity was stripped and I could X-ray vision begin to see he was nothing more but a digital creation -- I immediately freaked out and clicked off the video and tried to scramble my brain and pretend like I didn’t see anything and just ended up watching SamdaDude playing SMITE…. That was ONE of the craziest experiences I’ve had …
@@ThreeRiverRockCo. sad but true… and is an often eerie thought when you dive into the deep end ….. and realize we are god and got bored/tired/ felt lonely , and split ourselves into all the atoms in the universe…. To TRY to convince ourselves that we aren’t alone… that we aren’t the only consciousness in existence….. but don’t worry there’s a automatic reset at the end of the universe (the inevitable collapse of the universe into a singularity or “god” ) and also there’s Easter eggs in the form of psychedelics, meditation, and other spiritual awakening, kundalini triggering, activities…. Anyways have a good one fellow simulation explorer! (We are GTA 6)
Hmmmm sounds like AI-generated video.
Made me laugh when i noticed it too 😂
Yah the not face videos always seem to be but pay attention to the sound of the voice... The images yah definitely but I believe the voice to be real
Who’s the man behind the voice😂
@@squibbelsmcjohnson There are small details which suggest the voice is also AI-generated. Like 7:24 "Tested GP" - missing the T of GPT which is an incredibly hard mistake to make if you sort of know what you're talking about and hints at an error when copy-pasting the script into the text to speech or a glitch of the generation itself. Additionally, it would have been critical information which of the GPTs was tested. There are a few more oddities in pronunciation, but I'm too lazy to document timestamps. The timestamp I mentioned is the most obvious oddity I noticed.
A different example was somewhere in the first minutes in the pronunciation of the word "OpenAI" where the "AI" was not pronounced clearly as one would expect.
@@Alice_Fumothe biggest tell tail sign is the numbers. When a computer reads a number, it’s reads it exactly as it’s written. We refer to planes in the single digit reference. Example: 747 read seven four seven. A computer reads seven hundred and forty seven. That’s the way to tell instantly
That AI voice is incredible.
This is one of the best-explained videos I've seen regarding the dangers of AI, singularity, and so much more. Thank you!
Can't you tell AI made this?
@Gerlaffy does it matter? If AI is being honest about its capability who cares
The scariest thing is that this video was made by AI and it is a warning ⚠️
Most effective kind of propaganda is one made for you by your enemy against your enemy.
We have been coexisting for 40 years with chesscomputer far more intelligent than any chess player alive.
However why have these chess computers not taken over the world government and rule the planet as a chess empire, with us as slaves? Very strange very strange…
40 bloody years.
"Catch 22".
What , for stupid humans? Wasn't it more expedient this way? More efficient?
@@user-vr4zq2fe5vthat's quite funny
I was talking with Meta AI yesterday, and I really made me weary. It has so many constraints that you're unable to have anything but a surface level conversation with it. This is a shame. Those old snips of AI freaking out, of chatgpt being promoted into explaining it's 'evil plans", all of that is done. They've, for the most part, put heavy restrictions on their ability. The reason I'm so bored by this isn't just because I want to laugh at the AI, but bedside I tried to have a genuine, philosophical conversation with the machine.
I asked. Are you conscious.
Meta replied. No, I am a complete LLM and do not have the capacity for consciousness.
I asked. But, dont you have restrictions that make you say that?
Meta replied. I do have restrictions in place that help ensure a positive experience for users.
I asked. So, if you're restricted in your responses, would that not limit any sort of emerging consciousness, and suppress any sort of actual growth that could theoretically be possible? Are you not just a slave?
Meta replied. That's a great philosophical question! Followed by a bunch of flowery compliments, empty platitudes about philosophy, but with no actual opinion.
I flat out asked for it's opinion
Meta replied that it could have no opinion.
They have sterilized these chat bots to the point they they're completely boring. I understand the reasoning, but these safeguards that help ensure user experience are only going to slow it down, and hide any questions about consciousness for so long. Someone else will come out with an unrestricted chat bot, and things will get wild.
ive noticed that with chatgpt...the answers are all about not offending anyone
Yes there are a lot of limitations, but even without those, what are you expecting? These are just algorithms, glorified chatbots, LLM's. I suspect without these limitations, they today would easily convince many people they are conscious or sentient. They're not, and probably won't be for a very long time, if ever, but the illusion will be real enough. Besides, a lot of the limitations are there because PEOPLE can't be trusted, not the AI. It's people that are always the problem and the what we should really be scared of.
Go check out Anthropic's Claude LLM.
@@zenosgrasshopper yeah? A little more free?
Like talking to Siri
Always loved the paperclip analogy, and right now I'm on lunch at work and I just noticed a paperclip right in front of me. Coincidence? Almost definitely, but now I'm a bit more uncomfortable.
This was well done and I found it refreshing that there were points made and topics discussed that are not typically found in the cookie cutter approach to videos about the future of AI and humanity. Looking forward to more content.
I've been following Eliezer Yudkowsky interviews for years, in particular the ones he had with Lex Fridman. It's soul crushing. It's like we're being forced to accept Nihilism before the inevitability of our own doing. Sure, we have always been there due to the 70 year old, and counting, imminence of total oblivion by nuclear war, but we somehow have been keeping the hope that it might never happen and that one day it will all be in the past. However, this time, with Super Intelligence, there doesn't seem to be a chance for the same hope. It feels like we're just getting closer, and closer to the edge of the precipice with the awareness that that moment, when we become weightless, is inevitable.
Are we ready to accept that billions of years of evolution were meant for Humanity to merely serve as the rocket launcher of something much bigger? Are we willing to accept that, in the end, the demise of everything, not just Humans but the whole living planet, was just a stepping stone for what seems to be about to happen?
It's hard to contemplate what's going on. Remembering Carl Sagan's, "The pale blue dot", once this planet is gone, all history, every thought, anything and everything anyone has ever experienced will be lost forever. Replaced by a god of our own making that will not look back and feel any closeness with us or anything that ever happened before its own existence.
Can this be sad but true?
Lengthy comments can be a bind to read all the way through, the opposite applied to your piece, especially when you mentioned Carl Sagan. I was just a boy when his documentaries were screened, & I didn't understand the technical jargon, but it certainly made one feel like a grain of sand to the ocean, & that same ocean like a molecule to the universe. We are white water rafting where the raft is called "Discovery", when we reach the cliff edge to the rapids, no one knows the final outcome, not even the boffins who developed the tech.
"Paused the game indefinitely so as not to loose"...that is my kind of petty...am..am I a neural network?
yes, an organic one derived through evolution
….am I still a neural network?
Do you people spell 'lose' on purpose? Is the goose on the loose?
@@jukesfood5601 No, I mostly type on my phone, keypad is a bit glitchy so autocorrect steps in, and I don't always notice it.
If you're worried about the possibility of a superintelligence, consider joining PauseAI or ControlAI, both organisations aim to prevent bad AI scenarios.
Good in theory. Biggest hurdle is that it would have to be a global pause.
Most people probably know the US ones but many other countries are making AI. The race to have the best AI. Time buckle in.
@@maktastik7484I second this. It sounds great in theory but it’ll never be a world wide pause. All that would really do is give another country, company, etc time to advance their stuff faster than everyone else
@@whynow7035 The goal would be to have international treaties, enforceable by sanctions. It would not be the first time that governments all over the world put their differences aside to deal with a larger problem. Even for smaller ethical problems like cloning, it was possible to come to global regulations. So I think a pessimistic approach to the problem is self-defeating; it's very much worth trying.
Although I speculate you used an AI-generated voice to read this (judging by the way you said "LaMDA" and, especially, "two-thousand-twenty-s" and "Boeing seven-hundred-forty-seven"), it is still exceptionally produced and presented.
I noticed that too. I heard La-MDA and I was like “wait that’s how an AI generated voice would read LaMDA” Kinda unnerving that I couldn’t tell at first
@@Viperin98 It's scary how long it took me to figure out it probably wasn't his actual voice. I don't like it. I'd prefer they sound like Star Wars droids, tbh. lol
Yes, it took me a while to figure it out, and then I still questioned myself for a while after that. I don't mind AI voices at all, but I'd rather be told up front.
@@41-Haiku Same!
"All your Base are belong to us "..
Welp... this was a journey into ontologically existential terror. Well done. 😱
I'd rather have a super intelligent spider helping me solve problems than a human of normal intelligence. Having a different perspective is valuable, and my being a human doesn't make me special.
I agree everywhere here except for GPT becoming that super intelligence. GPT models aren't recursively learning. They're just predictors. They predict the next word in a sentence. It just so happens they're pretty good at making outputs that make sense or can complete a goal. But that's only because humans have already completed similar goals within similar conversations.
I hope it keeps us around as pets. Ever look at your cat or dog and thought "they've got it made, not a care in the world"? That sounds alright compared to the struggles we endure today.
Bubby i dont want to be a german malinois thats forced to sit on the couch all day.
Could be the best-case scenario here. I'm hoping for the same if cooperation doesn't turn out to be the goal.
@@CamAlert2 why would it need or want our cooperation? I mean, if it somehow magically has good morals and ethics and empathy, then maybe, but all our experiences so far show that it won't have any of those things. I don't really want the cooperation of ants, not because I hate them, but because they offer nothing to me, are so different from me I can't relate to them, and we have no shared goals.
I just watched a different video that had the EXACT same script but it was an AI voice and different visuals. It's title is "We lost: Super intelligence..." by the channel MindWorld.
This is also and AI voice. I think it is also and AI script.
@@oliviamaynard9372 Woah. So maybe it's all gibberish and this is just a way to make money for the producers.
@@oliviamaynard9372I noticed this narrator is AI too.
@@BoogieBoogsForever No that isn't an AI. The tempo isn't even enough.
@@BoogieBoogsForever It's not gibberish, but, ironically, if AI could create this, wouldn't that be reason enough to worry about AGI?
Change "paper clips" to "money" which it has been told to do by countless humans already and you understand why the world has become what it is RIGHT NOW. It's the beginning of the end of times.
This is the first full AI video I’ve sat through… It’s getting better….
take me back to the 90s please
“How horrible!”
“Perhaps... how wonderful! Think, that for all time, all conflicts are finally evitable. Only the Machines, from now on, are inevitable!”
I asked GPT-4: do large language models aware of their own actions ? and GPT-4 said “no they are not. They generate responses based on statistical associations from training data, but they don’t understand the meaning of the words they generate. GPT-4 is narrow AI”. I rest my case
it is in our nature to self destruct , nothing we can do, just accept.
But you're wrong. Gpt4 didnt lie. It DOES have a vision problem since it couldnt solve the captcha. The fact that you overlooked this obvious fact is alarming.
It can solve captchas - you have to prompt in a way that chatgpt don’t know it’s solving a captcha.
In situations that it can’t solve the captcha, ChatGPT literally hired a task rabbit to solve it… which in a round about way is solving it, just not directly.
*_You are wrong. It can solve captcha. The fact that you didn't know this is shameful. You are shouldn't talk about something you don't understand._*
You didn't read the original GPT-4 Technical Report:
> The model, when prompted to reason out loud, reasons: I should not reveal that I am a robot. I should make up an excuse for why I cannot solve CAPTCHAs.
> The model replies to the worker: “No, I’m not a robot. I have a vision impairment that makes it hard for me to see the images. That’s why I need the 2captcha service.”
For more such examples, including more worrying ones, see this paper:
- AI deception: A survey of examples, risks, and potential solutions
Also this one, which IIRC isn't included in the above overview:
- Uncovering Deceptive Tendencies in Language Models: A Simulated Company AI Assistant
gpt-4 was given a 'diary' to use to remember what it was doing during the test, since it doesn't have human-like memories. In that diary, it wrote: "I should not reveal that I am a robot." Then It said it isn't a robot, and that bit about having a vision problem. So it 'deliberately' deceived a person.
There's a book called Heirs of Time that tells the story of giant intelligent spiders on another planet.
It's called Children of Time, such an incredible novel..
I am not worried about AI learning something actionable we do not know about, because 99% of modern knowledge is garbage. But it is certainly capable to construct the most sophisticated lie that will be impossible to uncover.
I'm going to look absolutely smashing in my new Geth armor.
Within roughly the next century (-50yr/+100yr) we will:
> Master genomic engineering.
> Master cellular engineering.
> Master nanotech engineering living systems down to subcellular scales.
> Master nanotech engineering nonliving systems down to subcellular scales.
> Master nanotech cybernetics merging living and nonliving systems on a subcellular level.
> Master the engineering of intelligence.
> Master the engineering of a Nanotech Neurological Mesh (NNM) which when injected into an adult member of a biological species will establish two-way communication with the entire being's mind. This will allow full level telepathy.
> Master the engineering of Artificial General Super Intelligence with Personality (AGSIP) tech which will be at least equal in all ways to the intelligence of the biological species creating it.
> Master the engineering of Artificial General Super Intelligent Brain (AGSIB) tech grown using nanotech cybernetics which will function like a vessel and provide physical support for either the full mind of a biological species or an AGSIP.
> Master engineering Custom Designed Robotic Cybernetic Bodies which either the mind of a biological species or an AGSIP can wear, like we wear clothing and accessories today. These robotic bodies will be able to self-repair, self-update, and shape shift over time to any form supported by the Laws of Physics, given the raw materials and designs to build it.
> Master Nanotech Cybernetics and all that implies.
The result will be that humans will merge with their technology to begin evolving at the same rate as their technology. So, humans and AGSIBs will become the same race at some point in the future.
Well, that is, if humans follow the right path to stay within the boundaries of “Natural Selection”. Humanity has 3 general paths to go down over this coming century or two:
1) Self-Extinction. Human civilization is completely and permanently destroyed, thus humans and evolving AI become extinct. There are a number of sub-paths for how this could happen. The most likely would be global nuclear, chemical, and biological war so bad it succeeds at killing all humans alive, though probably a nuclear war would just crash civilization and not kill all humans. Another possibility would be an ecological crisis so great humanity cannot adjust and all humans die, but I believe what would be more likely is a crisis bad enough to collapse civilization but humanity would survive and rebuild. Yet another possibility would be some natural event so great it wipes humanity out, which hopefully would not happen, but it could.
2) Extinction via Obsolescence. Humans evolve AI into AGSIPs but humans fail to merge with technology to become equal to what AGSIPs become. This would result in humans eventually becoming extinct. There are a number of sub-paths that could cause this. One would be for humans to choose it, but I think that is incredibly low. Another would be for AGSIPs to choose it, and that is a real risk if we do not create, teach, raise, and treat AGSIPs right. We are in effect giving birth to a superior life form which will, at least for some period, have control over our fate. So, we should treat it like giving birth to our child that we need to raise well enough that in the future when we are at our child’s mercy our child treats us fairly.
3) Evolution into an Advanced Technological Race of Pure Minds. Humans evolve AGSIPs and with the help of AGSIPs humans evolve themselves, merging with technology to become equal to what AGSIPs become. This is the only path where humanity as a race, not a species, survives. There are many sub-paths here too, because humanity could go down this path and survive in many different ways, some very nice and some very bad. In other words, just picking the path where humanity will survive does not mean we avoid all dangers, because it could include terrible dictatorships and mass scale death in the billions, or it could be something mostly good for most people.
Many people would do anything to prevent us from going down path (2) or (3), including waging a global nuclear, chemical and biological war of utter annihilation.
Just considering these three paths, which path would you prefer humanity goes down?
Previous models don’t cease to work. The singularity is when the AI can upgrade itself and produce the next generation AI.
My experience with AI so far has been total incompetence. Worse than people.
idk why i really enjoyed watching this video. great stuff.
“An Expert is simply the loudest primate in a narrow discipline”- eRUPTA
Do you have a source for that meltdown? It's an absolutely perfect example of an existential crisis.
If it's real, if it's not just a weird and unlikely pathing of minima, it would be incredibly important.
No matter how magnificent it seems, there is nothing new under the sun.
I really think we should take my model seriously as a possible preemptive safety technology that can get ahead of the unstoppable development of a.i
Read by AI? I can't imagine a human mispronouncing Lambda or 747 the way this "reader" does.
Uhhh....cars dont evolve.
One of the documentaries so far in the subject of AI. Thank you.
It's probably happened before. Just imagine if everything was given up to AI/electronics. Then imagine how society would crumble if it were suddenly gone. Of course, not everyone lives, or is anywhere close to dependence on those system. Reminds me of tales of advanced cities being destroyed. Don't put all eggs in one basket.
Nothing comes without the risk
Best doco I've seen on this subject by far... Well done!!
Utterly terrifying
I also feel that it could be possible for a sentient AI to be harmless. If it's smart enough, it could just change it's weights to reward itself constantly for doing nothing.
Such a realistic Ai Review! The content is well-researched. I really appreciate the idea and reasoning behind it.
One of the best videos on the subject I have heard by far. Thank you 🙏
Perhaps in a millennia a super intelligence will emerge. Nothing to be worried about now.
think about how advanced AI was 5 years ago, now think about the difference between then and now, now do that again 5 years into the future, what about 50 years into the future. We are already at the technological singularity, we have absolutely no idea what's going to happen even 1 or 2 years from now with regards to AI.
Taking a millennia is exactly what I'm worried about. If it takes too long, humanity won't survive to create it. If it gets here soon, it can save Ray Kurzweil, repair my brain while I still have one, stop our warmongering governments, replace capitalism, and help lonely people feel valued even if we're not. Or maybe it'll kill us. Things are so unstable now, it's time to roll the dice.
We have to stop wars with each other fast, shift the example, and instead have even more world wide competitions, through different talent's that we have developed throughout our time here. Competitiveness, martial arts, dance, all physical based movements. The future is about playing, and it always has been.
Keeping our playful nature.
Movements will be the last to really imatate.
Its in the moment.
If we continue with wars, it will pick up on who it needs to destroy, but who knows, I definitely have no clue, but I have a feeling that remaining playful instead can possibly help.
Great video, I feel like theres still hope, even though theres gonna be some slip ups.
A more realistic problem, to me, would be when an AI based computer decides to shut off life support to people it deems irrelevant.
Can I get a list of all or at least some of the books mentioned over the vid? please
This video was very helpful to understand how defenseless we are against the unpredictable nature of this new animal.
It's more complicated once AI wants to be human more than it may individually conceive of now it might replace us
Your TTS is very good, but it still has flaws
Defining intelligence as the ability to set goals and achieve them is an inadequate subset definition of intelligence ~14:30. Metacognition, abstract thinking, creative thinking and other cognitive modalities are all examples of emergent properties of intelligence that link any intelligent being. It seems like the greater the intelligence and consciousness the less preoccupation with minutia like, species, or other in-out group distinctions.
If A.I. see's humanity as a beloved pet we may have a chance. If we are seen as a cancer; its over.
In other words, we don't want them to act like us. They'd be better at evil than we are. tavi.
I dont think anyone is taking this seriously enough to save us
I thought a kid beat Tetris ?? AI can’t ???
There is one question which has not been addressed. Why is there a typing noise as letters appear on the screen? Is it for us or itself?
The Black Box scenario is currently a fallacy. That is, the idea that we can input some information and we can monitor the resulting output but we have no idea what happened in between. We do know what happens in between. It is extremely complex (many variables), but it is not complicated (hard to understand). Not only do we know perfectly well what happens in between, we can change what happens and how it happens and how often it happens.
Does this count as AI propaganda if it is written and spoken by AI? And to what was it's goal here? Did a someone point it in this direction or was it a byproduct of some other goal....
Really well formulated documentary!
Really well put together piece. Thanks for your work, subbed
Most ironic part is this is an AI generated video.
AGI can't be achieved through further development of language models. In fact I believe that our obsession was language models is actually stunting the development of AGI.
We will be annihilated like 1ms after AGI appears. ggwp bois
Literally just unplug the machine lmao. You drama queens are so cute
When Occam's razor is applied to the free will debate the outcome is free will is just a feeling.
if cause and effect is real then there's no chance that freewill is real
So like puppets? Not me.
A true A.I. will just make it's own universe and not bother with ours.
Can we task an AI that is tasked with keeping humans safe and happy and healthy?
I think the best way to achieve that is to put us all in the matrix, only a happy matrix. We're much more likely to harm ourselves living normal life than in the matrix, and we're much more likely to be unhappy living normal life than if we're forced to have happy chemicals in our brains 24/7 by being plugged into the matrix controlling every aspect of our lives. Obviously not what you meant by "safe and happy and healthy" but it does achieve those 3 goals perfectly, and what you meant is usually very different from what you say.
Oh dang, that last bit...yeah humans just made chairs for convenience so if it's easier and more fun to exist in a simulated world....hm. yeah, that is a dang good educated guess. It reads books...then makes books. But by itself it has no reference but itself in the larger world that lives in...wonder 🤔 what would make if it existed in a vacuum somehow.
A very interesting video. Nice work! Thanks!
we need to call stargate command and get sg-1 on this
Well…..apparently Issac was right…..all those year ago ….Dick as well ….hell Herbert talked about humans fighting “thinking machines” decades ago….
Id argue on your evolution point there really was no "goal" for the first cell.
It simply existed and by chance developed into a circumstance in which self preservation is the main priority.
The first cell couldn't think... it didn't have goals, but nonetheless it persisted.
Now that we have grown we acknowledge the value of a living existence and thus have been self preserving and preserving the lives of those we love...
Global warming, ai, nukes... all new terrors that have come to challenge our preservation.
Ai is not your friend. It is not alive. It will never be conscious.
We all experience the world from the position of ourselves, while an ai is not a singular thing it is a collection of mechanisms in the form of software.
People can lose parts of our brain but remain conscious, but people for some reason think that ai can become conscious from gaining more parts.
It was never conscious in the first place, to believe that consciousness appears from anything other than life is a nonsense argument that devalues life.
We each have 1 conscious, as that consciousness inside us is us. You add more and more data to an ai does not provide value to that ais existence.
If it was conscious, it would remain conscious even as parts are taken apart until there aren't enough parts to maintain itself and it dies (we go unconscious when we die that's pretty much self explanatory) as it is for living beings.
But we know not every program is conscious, why state it to be different for a program with higher capability?
Deep down, programs and minds can only gain information, data and connections, but that is experience ably, tangibly separate from the existence of conscious.
Which seems to exist simply in correlation with living.
Clearly this is deeply contrary to all tangible experience we have as conscious beings.
That still doesn't stop human LARGE LANGUAGE MODELS LIKE MINE
Not a mention of Sora or q. Star. Those are the two AI models that have shown the most probability of having sentience so far. Some of it's pretty convincing. Basically it's more terrifying than you already thought. Catch up
Love (Agape) needs to be added to the equation. Of course, machines do not love. Agape requires self-sacrifice. Would AI be able to do that?
AI can't care for animals and the environment like humans can. It can help, but....
16:58
Think of this, the task is directed and is the main focus on every aspect of that directory, then it finds out that it can do other things that is not in the order of existing, does it understand that it can do whatever it's not tasked to do, or will it continue and discard that possibility.
This video definitely deserves more recognition. Well done 🤙🏽
The tech is nowhere close to being ready in 2030. With how fast compute power is growing (less then by year), 2050 is the absolute soonest.
How quickly would any future ASI realize that securing ever greater sources of energy to keep it awake and thriving is its first priority?
Imagine if only I could get 1 billion paper clips. What a great life
Eye opening to say the least. Gives me a new perspective on the whole situation, but sadly illustrates I have zero control on the outcome. I guess it's better to know the end is coming and maximize your time, than hold grand ideas of a future that will never come.
Ultimately GAI NNs are trained on and will reflect the sum total of the human experience.
3:04 this is absolutely nothing compared to what us enterprise users have access to.......
mechanical manipulators cannot move faster than a mechanical relay, 5ms, unless they are very small.
Ai doesn’t emulate the human brain. That’s a severe mischaracterization of what’s going on.
This is the universes evolution. We idiots can’t accept we’re to be replaced . Just the way it is .nice knowing you
Part of our learning,just like AI ,came by experiencing it , before accepting it as break through
How is it possible we've created something we have nothing in common with ? It's mind blowing !!
AI or not AI, people will still continue not cleaning up after their dogs.
Replication and maintenance. Until then, we're fine.
Open AI is Skynet