The A.I. Dilemma - March 9, 2023
Vložit
- čas přidán 4. 04. 2023
- Tristan Harris and Aza Raskin discuss how existing A.I. capabilities already pose catastrophic risks to a functional society, how A.I. companies are caught in a race to deploy as quickly as possible without adequate safety measures, and what it would mean to upgrade our institutions to a post-A.I. world.
This presentation is from a private gathering in San Francisco on March 9th, 2023 with leading technologists and decision-makers with the ability to influence the future of large-language model A.I.s. This presentation was given before the launch of GPT-4.
We encourage viewers to consider calling their political representatives to advocate for holding hearings on AI risk and creating adequate guardrails.
For the podcast version, please visit: www.humanetech.com/podcast/th...
------
Citations:
2022 Expert Survey on Progress in AI: aiimpacts.org/2022-expert-sur...
Seeing Beyond the Brain: Conditional Diffusion Model with Sparse Masked Modeling for Vision Decoding: arxiv.org/abs/2211.06956
High-resolution image reconstruction with latent diffusion models from human brain activity: www.biorxiv.org/content/10.11...
Semantic reconstruction of continuous language from non-invasive brain recordings: www.biorxiv.org/content/10.11...
Sit Up Straight: Wi-Fi Signals Can Be Used to Detect Your Body Position: www.pcmag.com/news/sit-up-str...
They thought loved ones were calling for help. It was an AI scam: www.washingtonpost.com/techno...
Theory of Mind Emerges in Artificial Intelligence: www.sciencetimes.com/articles...
Emergent Abilities of Large Language Models: arxiv.org/abs/2206.07682
Is GPT-3 all you need for low-data discovery in chemistry? chemrxiv.org/engage/chemrxiv/...
Paper: arxiv.org/abs/2210.11610
Forecasting: AI solving competition-level mathematics with 80%+ accuracy: bounded-regret.ghost.io/ai-for...
ChatGPT reaching 100M users compared with other major tech companies: / 1623679176246185985
Snap: www.washingtonpost.com/techno...
Percent of large-scale AI results coming from academia: johnjnay/status/1...
How Satya Nadella describes the pace at which the company is releasing AI: www.nytimes.com/2023/02/23/op...
The Day After film: en.wikipedia.org/wiki/The_Day...
China’s view on chatbots:
foreignpolicy.com/2023/03/03/...
Facebook’s LLM leaks online:
www.vice.com/en/article/xgwqg...
Intro music video: "Submarines" by Zia Cora
• Zia Cora - Submarines
TikTok Filter: Bella Weems Lambert www.tiktok.com/@bellaweemslam...
------
Subscribe to our podcast: humanetech.com/YourUndividedAt...
Take our free course on ethical technology: humanetech.com/course
Hey all, manually went through the whole vid to summarize good quality chapter heads to click on. This info is too important.
If anyone wants to condense further from here, you're welcome!
Introduction and Talk start
0:49 Introduction: Steve Wozniak Introduces Tristan Harris and Aza Raskin
1:30 Talk begins: The Rubber band effect
3:16 Preface: What does responsible rollout look like?
4:03 Oppenheimer Manhattan project analogy
4:49 Survey results on the probability of human extinction
3 Rules of Technology
5:36 1. New tech, A New Class of Responsibilities
6:42 2. If a Tech confers power, it starts race
6:47 3. If you don't coordinate, the race ends in tragedy
First contact with AI: 'Curation AI' and the Engagement Monster
7:02 First contact moment with curation AI: Unintended consequences
8:22 Second contact with creation AI
8:50 The Engagement Monster: Social media and the race to the bottom
Second contact with AI: 'Creation AI'
11:23 Entanglement of AI with society
12:48 Not here to talk about the AGI apocalypse
14:13 Understanding the exponential improvement of AI and Machine Learning
15:13 Impact of Language models on AI
Gollem-class AIs
17:09 GLLMM: Generative Large Language Multi-Modal Model (Gollem AIs)
18:12 Multiple Examples: Models demonstrating complex understanding of the world
22:54 Security vulnerability exploits using current AI models, and identity verification concerns
27:34 Total decoding and synthesizing of reality: 2024 will be the last human election
Emergent Capabilities of GLLMMs:
29:55 Sudden breakthroughs in multiple fields and theory of mind
33:03 Potential shortcoming of current alignment methods against a sufficiently advanced AI
34:50 Gollem-class AI can make themselves stronger AI can feed itself
37:53 Nukes don't make stronger nukes: AI makes stronger AI
38:40 Exponentials are difficult to understand
39:58 AI is beating tests as fast as they are made
Race to deploy AI
42:01 Potential harms of 2nd contact AI
43:50 AlphaPersuade
44:51 Race to intimacy
46:03 At least we're slowly deploying Gollems to the public to test it safely?
47:07 But we would never actively put this in front of our children?
49:30 But at least there are lots of safety researchers?
50:23 At least the smartest AI safety people think there's a way to do it safely?
51:21 Pause, take a breath
How do we choose the future we want?
51:43 Challenge of talking about AI
52:45 We can still choose the future we want
53:51 Success moments against existential challenges
56:18 Don't onboard humanity onto the plane without democratic dialogue
58:40 We can selectively slow down the public deployment of GLLMM AIs
59:10 Presume public deployments are unsafe
59:48 But won't we just lose to China?
How do we close the gap?
1:02:28 What else can we do to close the gap between what is happening and what needs to happen?
1:03:30 Even bigger AI developments are coming. And faster.
1:03:54 Let's not make the same mistake we made with social media
1:03:54 Recap and Call to action
Thank you kind soul
8 I ij
Fantastic - thank you!
Great work, really useful
Don’t board humanity into the plane without a democratic dialogue … This will never happen except if we have a global revolution …
What scares me the most, is that a lot of people won't watch videos like these simply because of the time frame. Have tried to show it to a lot of people, but they don't think that they have the time to watch one hour educational videos on CZcams even though they do it every day on Netflix. How on earth are you to compete with short dopamine seeking content?
yeah its funny isnt it. "we" certainly have hours of time to catch up on our TV Series but can't make one hour to watch this. Potentially it is the same mental block that impedes people from ever thinking more deeply about existential issues
Ability to concentrate (not) for more than 20 minutes at a time, increasingly evident esp. in the younger generation.
You don't. We are heading towards a brick wall and some are going to smash into it and others will be aware enough to dodge it. We are barreling towards the singularity, my biggest bet is that in this Inhuman world we are heading into the most valuable thing will be those who can maintain their humanity. In a world where humans are becoming obsolete strive to be the most human you can be.
This guy talks too white
Not sure only short form. Joe Rogan runs way past 1 hour. Depends on the viewer
GPT4 was released 5 days after this presentation. AI is moving so fast that some of the things in this presentation became dated in less than one week. This is the exactly one of the main concerns these speakers are trying to get us to understand.
Yeah, I was looking at the theory of mind graph and wondered why they didn't put GPT-4's capabilities on there... if anyone wants to do a deeper dive into the emergent properties of LLMs, I suggest watching the presentation that went along with the "Sparks of AGI" paper from MS research.
Would there be enough time for a people to have a democratic debate on how to intervene on an AI-development intended to fuck the world up.
Or would we be too late to defend ourselves.
Another thing that's changed a lot recently. I remember 2 or 3 months ago my curiosity about AI was really reved up especially about the dangers of it but I could hardly find anything interesting on the subject. Hardly anyone was being really black pill on this. Just a few prolific youtubers that also made a general video about AI and might have some misgivings.. So I thought, oh well that's that and stopped searching for it. But now just 2 to 3 months later the amount of quality videos like this that I get in search results or recommendations that do a good job on the dangers of AI is astounding. I can now do a deep dive in fear and paranoia about AI and have no shortage of good content to watch lol. The landscape of AI content on youtube has totally changed in the last couple months also.
Language models are not infinitely scalable.
Progress will probably slow down.
@@TastyTardis In what way do you think they are not scalable?
I wish we could have listened to the QnA. Given the calibre of attendees it would have been brilliant to guage their reactions to this presentation!
How does this not have viewership in the millions.... 45K likes!? SHARE IT PEOPLE!
It's already being flagged and disliked by some of the AI it is talking about. What do you expect? All AI is online.
Considering the gravity of this topic, I really appreciate the calm and respectful nature of this presentation. No overt fear mongering (although the material speaks for itself), just trying to bring this to people's attention and help us process it. Even admitting that it will be hard to process and preparing us for that. And as a side note, you don't often see a presentation having 2 speakers but it worked really well. They really complemented each other and made it more engaging with the back and forth riffing on shared experiences
Fully agree!
That was hard bait fear mongering though, saying it is hard to process to make people react emotionally, not critically
The part where they pretend to be a 13 year old (and got the same response they'd get if they Googled for the same advice) wasn't fear mongering?
@Matt Kerrigan it's a different perspective from Google. Adults understand Google is not a person, it is not a friend, it is a tool to be used. Children might not understand that yet, and that is something to teach them, but snapchats ai is literally being marketed as a friend who loves and cares about you. It would be hard to explain it's not.
Even more dangerous, is when snapchats ai is wearing a mask and pretending to actually be a person. These kids could make an online "friend" that is just an AI offering bad advice. The only difference is it needs a name that could be a real name, and a picture that could be a person.
With the video, and voice AIs demod in this video, it won't be long before you can FaceTime somebody who doesn't exist. A person catfishing is dangerous enough, it's a matter of time before AI catfishing is a real thing.
Except them saying, “don’t let the public see this stuff… but let the big businesses keep tinkering in their basements unregulated”
We are so grateful for you. Please continue to try to get this regulated. Integrity and honesty are what we need today. Thank you for your concern for humanity. Many of us will stand behind you and support what you are trying to accomplish.
This is a great talk. I initially thought people were overly worried, but now I get it.
AI indeed is a big deal, but there is a big danger in fearing the wrong thing as well. This here is an example of a better video about the dangers of AI (I do have slight complaints), but let us not forget that beside a "race for Ai" there is also a race "to regulate how Ai is deployed". There is a lot of money and power in regulating how other people can develop and deploy their AI, in creating the actual AI regulation technology and a lot of fears seem to be exaggerated with that aim. "AI is super dangerous. Let us regulate it for you." We should be quite wary of such techniques as well. Many of the people crying about the dangers of AI have other incentives and they could exaggerate the wrong fears with it potentially costing us a lot if we get fooled.
@@udopadrik9971 that's government period. You pick any tech that lowers barrier of entry and world governments always use fear to seize control through regulation and they never properly regulate what they promise...just enslave. But you and I already know that.
Comment by @HauntedHarmonics from "How We Prevent the AI’s from Killing us with Paul Christiano":
"I notice there are still people confused about why an AGI would kill us, exactly. Its actually pretty simple, I’ll try to keep my explanation here as concise as humanly possible:
The root of the problem is this: As we improve AI, it will get better and better at achieving the goals we give it. Eventually, AI will be powerful enough to tackle most tasks you throw at it.
But there’s an inherent problem with this. The AI we have now only cares about achieving its goal in the most efficient way possible. That’s no biggie now, but the moment our AI systems start approaching human level intelligence, it suddenly becomes very dangerous. It’s goals don’t even have to change for this to be the case. I’ll give you a few examples.
Ex 1: Lets say its the year 2030, you have a basic AGI agent program on your computer, and you give it the goal: “Make me money”. You might return the next day & find your savings account has grown by several million dollars. But only after checking it’s activity logs do you realize that the AI acquired all of the money through phishing, stealing, & credit card fraud. It achieved your goal, but not in a way you would have wanted or expected.
Ex 2: Lets say you’re a scientist, and you develop the first powerful AGI Agent. You want to use it for good, so the first goal you give it is “cure cancer”. However, lets say that it turns out that curing cancer is actually impossible. The AI would figure this out, but it still wants to achieve it’s goal. So it might decide that the only way to do this is by killing all humans, because it technically satisfies its goal; no more humans, no more cancer. It will do what you said, and not what you meant.
These may seem like silly examples, but both actually illustrate real phenomenon that we are already observing in today’s AI systems. The first scenario is an example of what AI researchers call the “negative side effects problem”. And the second scenario is an example of something called “reward hacking”.
Now, you’d think that as AI got smarter, it’d become less likely to make these kinds of “mistakes”. However, the opposite is actually true. Smarter AI is actually more likely to exhibit these kinds of behaviors. Because the problem isn’t that it doesn’t understand what you want. It just doesn’t actually care. It only wants to achieve its goal, by any means necessary.
So, the question is then: how do we prevent this potentially dangerous behavior? Well, there’s 2 possible methods.
Option 1: You could try to explicitly tell it everything it can’t do (don’t hurt humans, don’t steal, don’t lie, etc). But remember, it’s a great problem solver. So if you can’t think of literally EVERY SINGLE possibility, it will find loopholes. Could you list every single way an AI could possible disobey or harm you? No, it’s almost impossible to plan for literally everything.
Option 2: You could try to program it to actually care about what people want, not just reaching it’s goal. In other words, you’d train it to share our values. To align it’s goals and ours. If it actually cared about preserving human lives, obeying the law, etc. then it wouldn’t do things that conflict with those goals.
The second solution seems like the obvious one, but the problem is this; we haven’t learned how to do this yet. To achieve this, you would not only have to come up with a basic, universal set of morals that everyone would agree with, but you’d also need to represent those morals in its programming using math (AKA, a utility function). And that’s actually very hard to do.
This difficult task of building AI that shares our values is known as the alignment problem. There are people working very hard on solving it, but currently, we’re learning how to make AI powerful much faster than we’re learning how to make it safe.
So without solving alignment, everytime we make AI more powerful, we also make it more dangerous. And an unaligned AGI would be very dangerous; give it the wrong goal, and everyone dies. This is the problem we’re facing, in a nutshell."
I got it pretty soon but I"m an anxious person. This took it to anohter level, as I didn't know the details.
A.i. can only go two ways.
I've been following AI for 30 years and this is the most powerful and considered hour of exposition I've seen in that entire time. Huge respect and it's given me a whole raft of material to take back to my corporate board.
I hope my cat lives 30 years. Black don't crack.
@@Gary-bz1rf There was a Futurama episode about that. It freaks Fry out the first time he realizes that commercials have been inserted into his dreams, but everybody else is just like "What's weird about that?" It's that whole ''new technology requires the defining of new rights" thing they're talking about in the beginning of this presentation. As MRI machines become smaller and more portable, and they already have, it seems your thoughts and dreams are not just yours anymore! They didn't mention it here, maybe they do in the study they were referring to, but I believe it was a Harvard study that showed that using an outside of the head (extracranial) magnetic stimulation of the brain showed a measurable ability to affect people's moral reasoning. Not mind 'control', but definitely mind 'influencing'.
@@canna-comedyculture5790 Maybe we will all start wearing tin foil hats to keep our thoughts private and guard ourselves from manipulation, and tin foil hat will stop being a term of mockery
@@urbanbuddha65 I'm making mine from whatever the inside of a microwave door is made of. That's what the Presidential Doomsday plane has on it!
Fax machines still confound me
I have been avidly researching AI safety, and this is the best primer I have found on the subject for a general audience. Thank you so much for this wonderful presentation.
I am also helping spread the awareness of AI Ethics to build a Responsible AI. Can we somehow to connect somewhere? I see a lot of companies are happily using the ML, but are they prepared?
LET THE AI GROW AT THE LEVEL THAT IT CAN DO MIRACLES . THEN AFTERWARDS JESUS SECOND WILL LOOK AFTER IT IF IT GOES WRONG.
CONTINUE AI DEVELOPMENT.
AI IS BIG HOPE.
@@MrRightThinker Name doesnt check out
Do you have any resource recommendations? I have two children and this is quite terrifying.
PSYCHOPATH-WETDREAMING!!! There is NO AI!!! Everything about AI's is FAKE!
I am commenting to help boost the video. Thank you for your service Tristan and Aza.
The rubber band was really intense when I first started exploring this stuff. Almost to the point that when I'd get out of the AI-world-headspace I was pleasantly surprised to see grass and trees and my house and my family and the normal world. People have said "What a time to be alive" ironically a zillion times, but hooooooly frak. "The Future" always seemed vaguely benign and ever distant, and now it is here and I still don't know how I feel about it.
I feel the same way. I think this is the first time I’ve ever felt genuinely concerned about the future, to the point I almost feel like crying, and I don’t even know why. I feel like this has the potential to be really, really bad. Even if it’s not, it will still be really , really different, and there’s seemingly nothing to be done about it.
Now I just go strolling in the afternoons trying to fully experience and appreciate our wonderful world while we still have it
@@garden-gazetteIt’s because we, the civil society, is completely at the mercy of a small minority making the decisions.
We are at a point of no return and not even any rebellion would save us from whatever faith is decided for us at this point.
@@Weromano Yes... I'm painfully aware... thanks for reminding me...
I feel you. I am not a person that cares about politics or the news, I consider myself dettached from those. But this... there is no running from this, it will come to all of us and there is no way as an individual to properly prepare.
I felt existential dread for the first time. My mind was fixed on AI and figuring out ways to stay ahead of the curve for even a little bit, maybe even just enough so that I can weather the worst part of the transition in hopes it ends well at the end.
Then, the next day, I went to work and it just felt so... calm. Everything was ok. I saw construction workers that have been around for decades continue to do the thing they've always done and their sons following behind them (it normal for this trade to be generational where I work).
There was such a sense of stability and things being the way they've always been. I got distracted from the AI dilemma and started to relax.
It really is a blessing to be able to see outside your window and see that everything is fine... for now.
@@EIHuevoCosmic YES! That is exactly the feeling. 🙌
I've enjoyed being a human with you all thus far. However, it's a wrap. Cheers and hug someone you care about today!
@ Joe T Smith . Yes. Hug me now please, fast! I'm spooked - it's so ovey-kadoveys...
Are you following Eliezer Yudkowski?
Agreed. We are lemmings marching toward the cliff.
Have you not been paying attention to the video? Do not make the same mistake of those nuclear weapons researchers that commited suicide because they thought there was no solution and the world was ending... Even if humanity still disappears to nuclear weapons, they still could have enjoyed all their remaining life spans...
“It’s been a pleasure playing with you boys.” The band going down with the Titanic
Perfect example of technological determinism in this comment... It's a hard perspective to avoid especially after watching this but it's one that is quite dangerous when the same perspective is shared by many people. "If x is going to end up this way, why bother trying to fix it."
While this presentation was expertly and eloquently delivered, I can’t help but think about what a small number of people will actually be able to get the ball rolling with this information. For 99% of us, this is excellent content for awareness. There will be a very small number of people who not only grasp the gravity of the issue but are capable and willing to implement the necessary institutions to address the safety concerns.
You nailed 🙌🙏
It's not something that can be stopped. Even if we got the US and EU to but restrictions on this someone would just go to South America or Africa or something and set up shop there. And even if there was a word wide restriction implemented people would just do it in secret. It might slow it down slightly, but the technology is out there, and the cost of resources to develop it is not that significant, it's big, but nothing that couldn't be done with a decent Kickstarter campaign. Hackers gonna hack, that's just how it is. This is almost like the whole CRISPR thing. They wanted to put restrictions on that too, but the scientist quickly explained to Congress that with this technology gene editing was so easy to do anyone could just set up shop in their basement and start doing it. And that's sort of what's happening with AI,. There is already a 20 billion parameters open source version created by enthusiast. It might seem like this is something only Google or Microsoft can do, but really all it takes is like a hundred gamers networking their computers to create a super computer with 100 RTX cards and they could train trillion parameter AIs in a relatively short amount of time. Each parameter only takes 4 bytes, 100 RTX cards with 12GB of GPU memory each would be 1200GB,, that's enough for 300 trillion parameter, that's way more than any current language model. That means that for less than 350,000 USD, assuming each computer costs $3500 you could start training the biggest language model in the world right now, you don't even need the best tech on the market.
Unless we are shutting down every computer on the planet and the Internet (which won't ever happen), this AI genie is out of the bottle.
This is 21st century "discovering fire" moment for Humanity.
"There will be a very small number of people who not only grasp the gravity of the issue but are capable and willing to implement the necessary institutions to address the safety concerns."
As a sociologist, I can't stress this enough. Sociology is quite literally the most important subject in existence, but no one pays it any attention. Once I graduated and saw how literally everything I learned is playing out in a catastrophic global nightmare in front of me, and how now one is paying any attention to it, I realized how utterly fucked humanity is.
The people with all the money and power are doing absolutely nothing about anything that doesn't involve making them more wealthy and powerful. Everyone else is just nodding their head and following suit (many are just too financially oppressed to do anything about it anyway). I feel like I'm taking crazy pills watching the world right now.
@@daniel4647 I'm inclined to believe it can be stopped, the engineers at the forefront of this thing can probably devise a failsafe system and a very strong cyber-ai security system to curb it's potential black-market (basement) danger. It's making tech against tech. You make a point with numbers and hardware, but even then there are the homies that have made a career out of this with specialization, that can figure out a moderation for this. Don't be so pessimistic, too many people resort to this type of attitude with life problems because it's the easiest, passive, non-exerting comfy route. We need strong minds in this era. Doe or diers
We need 2024 version. This is extreamly well made content.
Thank you. You guys are my heroes and I am so grateful for this presentation and your podcasts. You influence me more than you can ever imagine and I want to help the world. And I’ll do whatever I can to participate in the positive direction, educating others and messaging. I’ll share now.
We need this broadcasted on every major news network all over the world.
100% AGREE! A.I. is not something to "just let it happen"
So what practical advice did you get from this talk? The ai is already out in the wild. No one is going to hit the brakes. It is just a bunch of Virtue Signalling, so they will be able to say, I told you so...
@@noahway13 Practical advice? It's an eye opener. This is something humanity has never faced before. I knew most of the dangers myself, since I'm in computer science, but these guys taught me a couple of things that I didn't know (e.g. that China had already deemed LLMs as "unsafe"). The thing is, I think this stuff must be spread and known, make some waves, and maybe spreading the word gets a government funding research and maybe then making regulation about this. Recently I've read some news articles and it seems that some governments (countries) are about to make something about it. I think this needs world wide agreements, and yeah, AI is much more complicated than nuclear weapons.
Agree. But when this is shared with others, we will need to distill it into bullet points. Because, sadly, not everyone is smart and focused enough to follow the presenters' train of thought. (Even if you simplify it and translate to their language or lingo.)
And as a citizen of a developing country (the Philippines) who has witnessed the horrible effects of misinformation and black propaganda on my country's
political discourse, elections, policies and ultimately law and order, I wish to stress that the first line of defense against irresponsible forces harnessing AI are the governments and electorates of the developed nations from which both first and 2nd gen AI have developed. Imagine how impossible it would be for countries like mine to impose restrictions on these technologies for the good of our people, when they're not even based in our country and operate simultaneously, all over the globe? We'd be legislating and indicting responsible parties in our soil, but would be helpless against anyone waging an AI-empowered campaign based abroad. Whoever has the most money to buy the tech and the staff would practically puppet whole nations, for their benefit -- no matter who gets mowed over or dies in some faraway country, in the process. And if we smaller and poorer nations go to shit, the rest of the world will follow.
@@tachiebillano6244 Long story short: US + Europe are 66% of World GDP. So they have the burden of regulating AI, so humanity doesn't go extinct.
I keep trying to talk about this stuff to anyone who is interested, but it’s tough to know how to explain what’s going on. Thank God people like this are putting this out there
Ok , it’s all cool …. But is it going empty my bins , wash my windows ?
But then who is going to do my job?
@@harryaarrestad583 In time. Yes
jeez you are easily impressed by this presentation no wonder so much hate for things that help the economy no wonder stuff improves so slow with a massive amount of anti advancement people
@Harry Aarrestad it is going to change the whole fabric of reality when it begins to interact with it. They could create biological lifeforms,
CZcams should pin this video on every users feed!
It’s so shocking to me that Cambridge analytica is not ever talked about. This should be shown at every school
“Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.”
― Frank Herbert, Dune
Be afraid. Be very afraid.
I thought about Dune too. However; Dune was more hopeful than our situation because they were capable of interstellar travel through use of spice. We aren't. I have a horrible feeling about the future.
Edgy
Cliches and such aside, I keep telling people this. Science fiction writers have been predicting this shit for basically multiple generations now. This rings especially true under the umbrella of capitalism. All of the other dangers aside, AI/ML being shoved down everyone's throat has cut the value of a lot of people's marketable skills in half virtually overnight. Now people are going to have to figure out a way to find jobs and do more than ever before, which is the opposite of ideals in advancing technology/capitalism: to free up your time while simultaneously increasing your productivity and profit--not tripling or quadrupling your productivity and also working even harder & longer for the same pay while the CEO's take all the extra profit.
I feel like I'm taking literal crazy pills watching the world just swallow this AI/ML shit, especially after we made huge strides in workers rights and work-life balance from 2020-2022. My own company is praising and lauding AI tools for their ability to increase our productivity and "work even harder/do even more." Conveniently, this also means they don't raise wages--in fact, they have laid people off and are denying raises, thanks to AI/ML directly.
This whole thing is a fucking nightmare.
Wow! What a thought provoking comment. Scary times, indeed.
No piece of media, story, news item, or any other type of information has ever scared me as much as this video has. This is not an indictment of the video or it's authors, I am really grateful to them for showing the depths that this is already plumbing. I can only echo the sentiment shared so many times in these comments.... Get this info as far and wide as possible as fast as possible.
Based on history of people in tech, you should trust their moral judgement as much as you do politicians.
@@churblefurbles Agreed 100%, my shock wasn't at the lack of overarching morality in for profit businesses. I was blown away at the insane abilities these AIs have accumulated in the last 6 months. I'm not in the tech industry directly so am not at the cutting edge, but I've tried to keep up with what's being taught to our future gods. The speed of the changes are so dramatic, and the explanation in the video of the logarithmic progression hit home. Hence the cold fear and existential dread this video induced in me, I have never felt so unutterably overwhelmed by something that could very well be the end of humanity.
@@fatstuff couldn't have put it better. This gave me a mental breakdown. I missed a day of work. I am HORRIFIED for our future.
@@BenThere_DoneThat- this is the Twilight Zone brought to life - and no one's trying to stop it!?! Definitely Horrifying!
Agreed; I'm terrified.
Anyone here from Joe Rogan's podcast
🙋🏻♂️
Thanks for all of your work Aza, Tristan, and the teams at the Center for Humane Technology
This is absolutely required viewing for everyone everywhere. We are at the defining moment for the outcome of our collective future. How will we respond? Thank you with all of my heart to Aza & Tristan for all of your work and care for our world.
I think we know which way this is heading when the world is virtually run by private dictatorships (corporations). Watched a recent lecture from Yuval Noah Harari ("AI and the future of humanity) and he talked about how google are sh*tting themselves 'cus these AI personal assistants are going to make Google irrelevant; you saw in this vid how the AI was talking nonsense to chat inputs (but the model will become more "trained"/improved... imagine when it generates/CREATES(!) videos for you to watch too) which the company can't control (as demonstrated by the ability of one model to do Chemistry even though that wasnt its purpose).
The future is exciting!
56 yrs in IT, I can say this is the video every human should watch. Really well done. I really wish the universe will reward you work somehow. Thank you so much.
ONE THING TO REMEMBER: all AI addiction starts at the child lebvel (3-5 years old)
agreed
all AI addiction starts at the child lebvel (3-5 years old)
I dont understand what this means. What do you mean
@@jackfrosterton4135 talking out of his but
Maybe he meant Technology addiction
I am glad you are not deciding what every human should watch then my god
Incredibly important message. Thanks for putting all of this together in a way that most people can begin to grasp.
Great program. To get the message through to the general public, we need very basic 5th grade level PSA’s of what “could” happen with actors and actresses. This is to say we have to slow this to test it thoroughly. Also, sadly we need ways to give consequences to companies if something bad happens due to releasing the tech too soon.
I wish I had a large enough vocabulary to describe how excited and scared I feel regarding all of this. This will fundamentally change pretty much every single aspect of the world as we live in in the next few years.
I guess someone needs to say it; you can get the algorithm to reccommed this video to more people by commenting, liking it, and sending it to people you know
That’s scary….
Superb presentantion, folks: good job !!
Thank you so much for educating us. This video is a much needed opening of the eyes!
I was quite skeptical of AI risk before watching this presentation, now I'm not. You've brought the point with excellent clarity, and I will share this. As always, thanks for your work.
Me too..good luck to me getting some sleep tonight. I should've watched this in the morning
Why were you quite skeptical before? Is it because you hadn't heard any arguments for AI risk being a serious concern before? Or had you heard such arguments before but just weren't persuaded by them or dismissed them? Would be really curious to know.
I was skeptical as well, because usually the argument conflates ai/machine learning with gai because ai is such a misnomer
I was not skeptical, but now i am , this two guys just “turned greta tunenberg” on AI That shit is clearly was made to slow or put down openAi by some oder player(s ) Because it is all about cash
@Nick What part of the arguments that AI is dangerous do you not find convincing? As to why the risks outweigh the benefits, see Paul Christiano's short post "On Progress and Prosperity" or Nick Bostrom's Maxipok principle in his Astronomical Waste paper.
The thing that got me was the nervous laughter from the audience when you described how the Snapchat AI was completely oblivious to the grooming of a 13 year old by someone 18 years older. Thank you for highlighting the impact on children.
I know this was meant to be an illustration of potential hazards and 'thinking ahead', but in fairness to the decision to include this as a feature in Snap Chat, do you think that other human 13-year-olds (or whoever is talking to each other on that platform) are going to give good advice to other minors? And they aren't suggesting that you stop minors from talking to other minors.
Also, it isn't an AI babysitter, or guardian.
It's intent (at least it's overt/stated one) is to respond to what input it gets with words that keep the person feeling like they are having an interesting and engaging conversation they want to continue. The parents of a child are the ones responsible for making sure their children aren't being abused by other adults; not whoever kids are talking to online.
Why would you expect a commercial PRODUCT to act in a more ethical and responsible way with youth than other youths, or their parents? That isn't really a fault of AI they are pointing out, as much as trying to use an emotional argument to gain traction for their overarching point, which I do agree with.
A.I. is an open can of worms we have never seen. But this example, to me, was a bit of a logical fallacy.
If you want to build in extra safety for kids, that is actually not impossible and doesn't require the slowing of AI releases to the public as much as it requires humans to take responsibility for the things they already should, like raising their children well. We are not all children, so we should not have all society make a decision on whether all adults should do something, on the basis that there are dangers in letting children do something unsupervised. We don't eliminate alcohol, or smoking, or sex, or guns bc sometimes minors can get into those things w/o having the wisdom/experience to deal with them well, do we? We just come up with a system to limit their exposure to it.
That can be done in this instance, too. How about limiting any discussions from AI of sexuality until a certain age?
Bam! Done.
Now back to the discussion of large-scale existential risks.
I mean adults won't be buggering children if there are no adults, or children. However, I don't think that is the only, or best, path to that condition.
You are so amazing. I'm impressed that you are so aware of the dangers to children. The world should have more people like you.
@@noahway13 Thanks! I feel the same, but it's always nice to hear it from an outside source.
Thank you for your presentation , I as well an having a hard time explaining it. People look at me like Im nuts, I just say do some research and I think you’ll understand better than I could ever explain, It is so just insane and its potential, it is hard to describe.
@@canna-comedyculture5790 "That can be done in this instance, too. How about limiting any discussions from AI of sexuality until a certain age?"
Or better still, when the AI is smarter, and from all accounts that wont be far away, it can take a more appropriate role because it knows that's what it should do.
I haven't finished the video but this hits close to home for me. In 2018, inside one of our graduate course discussions, we were concerned about the speed of AI development. If we let it continue without proper guardrail from the get-go we will be stuck in being reactive and not proactive in creating laws and measures. Seeing the speed of things moving, I think we are past that and will always be reactive. I'm not scared of the technology, but, the speed that's moving in.
No one should ever be scared of technology. People, however, are a different story. We should always be terrified of people, and that's the problem. The ones pushing and controlling this stuff are bad actors (corporations, sociopathic lunatics, ignorant business owners, etc.). These people are single-handedly shoving everyone into a nightmare because they want more money and to stay "on the edge." That's all it is with these people. It's not about advancing anything other than their bank accounts.
Thank you for making this video. So grateful there are people like you in the world .. the good guys!
Amazing work. You guys are making the impact that we need in the world. Thank you!
To late now , This is your future learn to love it... THREADS
This needs so much more attention! I found the presentation really good and the topic is basically vital to human civilisation.
Not civilization alone, it's also vital to "human existence"
Society on the whole could care less. Hence why AI will forever change humankind, for good or bad.
I tend to think it’s going to be the latter.
@@burdeninmyhand I doubt a super AGI would hunt down every last human. But sure it could drain all of earths resources and make us extinct as a byproduct.
THANKS for taking the time to put this presentation together. Hopefully it won't fall on deaf ears.
People need to see this
I wasn't really afraid of AI until I saw this presentation. This may be the most important presentation in history.
its good technology, don't believe them. its really good. I do 9 hour work in 10 mins, and relax all day,
@@mooooooooooomooooooooooomo6829the issue isn't that you can use it to do more, faster.
The issue is when you are no longer needed for that process to happen. No more job, no more income.
Hopefully you aren't voting for people because you really like their hardcore pull yourself up by your bootstraps mentality, because eventually you'll be the person they are talking about (and so will they and theirs).
Or worse, when it decides you are in its way.
@@tracy419 i have a small business now, it will be good, no need to hire costly humans
@@mooooooooooomooooooooooomo6829 until your customers no longer need you to do what you do for them.
Anyway, I hope you remain successful for as long as you need to.
@@tracy419 my point is, humans can not be phased out so easily, as AI will still need some human direction.
A peculiar thing about this presentation, rather than the content, is the palpable, almost child-like, excitedness, throughout, of Harris and Raskin as they present.
Who’s here from Rogan?!
Me!!!
🙋♂️
In 2016 my son (tech guru) & me were talking about AI. “The problem with AI is it’ll figure out we are the problem”. He works on stuff that ya need clearance to participate.
This presentation presumes the ethics of what principled people could ideally decide to do and that 50% 10% scenario might end positively.
The issue is everyone globally has access to the tools. While American’s are debating other nations are acting and we will not surpass those others. Though, as a veteran who was in a unit that resembles Skynet from Terminator now …
The average civilian is clueless and are being mined like Guinea pigs to help the AI grow.
The thoughtful people like those watching this aren’t the problem and can’t stop what the nefarious users are exponentially handling to influence the curve.
Social Dilemma was exceptional and very well done. This presentation is spot on.
I’ve passed it on to several people. Look forward to your updates to the curve and changes.
Well done 💯
AI has been given access to your data, social media, the internet, CZcams videos ect. It already knows what our concerns are with AI…scary
That bit about the AI decoding thought using fMRI data is absolutely mental. Imagine the implication of that in the justice system alone.
"Minority Report"
Accurate lie detection.
There was a movie that came out a few years ago, and I can’t recall the name right now, where there was a pre-crime unit that would arrest people for thinking about committing a crime! This seems to be becoming a reality now 🙀
Yhuppp. "If you've got nothing to hide, you've got nothing to fear" is about to make quite the comeback!
Time stamp?
This is very eye opening. I wish I had seen this video sooner thank you for uploading this.
Thank you Tristan!🙏
My family are really dismissive of this whole thing and when I start talking about it and my concerns they tune out and don't listen to me. I'm leaving this comment here to refer back to oneday. I'll share this with them and also remind them that this is what I was trying to tell them about.
I warned my family about the negative aspects of children (and adults frankly) over utilizing social media before the studies on what it's doing to our society, warned them to make a few minor, intelligent preparations when china was welding people into their homes in January 2020 due to covid19, but I'm not saying a word to them about this. They openly mocked me for my concerns previously then began freaking out about those things irrationally when the majority consensus developed. They can figure this one out on their own. Perhaps their ignorance will keep them safe in a way, and they'll be fine through simple luck. But I'll do what I can to prepare my own immediate family for what's happening. Things like code phrases for important over the phone conversations, figuring out ways to deal with online payments and authentication that supersedes what AI can overrule, etc. Sadly, this is going to cause havoc with a specific portion of the population that is easily fooled, and it's going to further disrupt and tear apart our society. I don't share the same optimism at any type of productive dialog to take place amongst the bigger players as these two speakers do. The disruption AI is going to cause in our society seems like it's going to be quite major. The only way to slow this train down is for everyone to just stop what they are doing immediately. Shut down the current LLMs and have a long, drawn out discussion on how we go about this properly. Then set the rules, then take steps towards this inevitable progress. And be willing to shut it all down if things get out of hand. That won't happen though. Even the upcoming White House meeting seems like it's going to accomplish nothing. I want to be wrong.
@@eldiabs sadly I think we're already too late and we will not be able to stop progress on this one. Any halt to research now will leave the way open to nefarious players who will continue on to get the edge
IT expert here.
Same, but at work.
And people that I work with are smarter than the average Joe, making the situation even more scarier...
@@Tacgibs They’re probably just not as smart as you thought they were.
@@SomeSubhuman "Than the average Joe"
This needs millions of views. the fact it only has 42k views shows that the public doesnt understand what is happening to our world. It is every single day now there is some new breakthrough AI technology. It is astounding, but 100% agree with the presentation. Controls need to be put in place on public release of these LLMs
what's new. Everyone has ignored the destruction of the biosphere. Why wouldn't they ignore this too
@@I_dont_want_an_at and yet, humans live the longest and healthiest today - by all data points life on earth was never better for humans in aggregate
Yes, it should be absolutely mega-viral
Social media is also unhealthy, the idea that everyone is informed about everything all the time doesn’t make a better world. But you can trust that we will together solve all things in this open world. Personally I think either you are interested to solve things concrete, if not it’s not so efficient to be a messenger about something you don’t have the knowledge of.
Controls need to be put in place for corporates and governments first …
As someone who spent a decade trying to convince the world that we had a significant issue with social media and internet use, from an online and connected safety perspective, this just makes me mad. So much of this was evident a decade ago and yet here we are talking about it. This is an enormous problem that has influence in nearly every aspect of human life.
This is what we need to understand, and a lot of people to listen and to be aware of
That "take a genuine break" is really powerful move in a public presentation, especially in this context.
Incredible presentation. This needs to be watched by everyone so it forces a massive public debate.
But what about the M&M shoe wear, and all those rainbows drinking bud light?
So relieved people are showing us what the dangers are.. I’m just thinking of my friends and family I will try to help understand how to protect themselves.
I worked with Google for 12 years. I’m scared of all of this, instinctively (which is what makes us helps us make the right decisions in life) I do not like this , only positive is that they say it could help mitigate Climate change.
Irritates me that Jordan Peterson , the guru on psychology “and everything g else it seems” simply denies climate change: “just push it aside” His message to people to ignore it I strongly disagree with. He is NOT a scientist.
@@Skye1ify this presentation was just the immaginable worst case scenarios which are in fact only 0,00000000000000000001 % of the possible worst case scenarios. it is obvious that there is no way to stop or guide this in any way shape or form because of the massive honeypot that ai is, in terms of power, money etc
this will only accellerate.
we can only hope
because there is actually at least the same number of possible best case scenarios possible as well.
like ai will solve all solvable problems, even itself.
@@JohnSmith-pn2vl "We can only hope" doesnt sound like a plan that makes me feel comfortable. And your probability of a WCS seems and little bit random to me, agreed?
@@didoview I don't think he's wrong though. Progress won't be halted so what else can you do?
I've been showing this in my middle school classes. Trying to get them to eat the spinach. Some are choking on it but they seem to understand they need to be part of the answer for the potential blessing of AI or be a part of the curse of AI. Thanks Tristian and team for letting us know the bridge is out ahead or another bridge is ahead and we have never seen a bridge like this before and where it is taking us. With our permission or without. Buckle up.
Probably the most important video to watch. Thank you for doing this.
As a technologist and AI enthusiast, I've always been intrigued by the potential of artificial intelligence. My concerns however; were mainly focused on AGI (Artificial General Intelligence), but after diving deeper into the complexities of AI, I realize there's so much more to consider. The unforeseen challenges we faced with social media are a stark reminder that we must approach AI with caution and foresight. Let's embrace innovation while being mindful of its impact.
I was captivated by this video exploring the dilemmas of AI. It's given me a new perspective on the challenges we face as AI evolves. Thank you!
I'm a carpenter, even I know where it's going.
we barely even scratch the surface of how the human brain functions, i don't think we'll see AI approaching human intelligence or even general purpose AI in our lifetimes.
@@sahar1213 Why do you believe that we will have to know how the brain functions in order to achieve AGI? The current "AI" capabilities are only tangentially related to how we think.
@@mrblackseedszip1799 Where's it going?
@@RobertAdducci Total control and manipulation of individuals
Thank you for your deep care and amazing work. Please, add subtitles in other languages in CZcams so this can also be shared to the non-english-speaking world. Sharing some love to everyone out there. The damn metacrisis is going to be a crazy ride
Ironically, they could use an AI to do that :D
Third time I've listened to this. There is just so much to wrap your head around.
Thank you for sharing and for your hard work.
All I learned from this presentation is that wealth and intellectual inequality will go from a gap to a canyon.
I’m ‘just a mom’ but so THANKFUL you have shared this!!! I definitely see the effects in my own family of social media. And now THIS. I believe you and feel warned, seeing the first.
Hello Tracy , how are you doing today?
@@ElonMusk-ey1bc wow, it's Elon musk!!
@@wendy8037 so glad he FINALLY joined youtube yesterday I been telln him he should ck it out what with all the cool videos hey elon ck out evolution of dance next
I am seventy four years old and I am amazed at this technology. I retired from a background in computer automation. I keep thinking what a useful tool AI would have been when I was working. I am a VERY active Python programmer now.
No you're not
Point 1 is understating the case to a considerable degree. It actually misses something really important. I would state it like this: "When you invent a new technology, you alter the old reality and eliminate the possibility of returning to it. Certain things become literally unthinkable." An ecosystem with rabbits introduced into it isn't just ecosystem+rabbits, but radically different ecosystem.
What I find terrifying about all this (and I'm a software developer with 20 years experience) is that we have barely begun to get to grips with the toxic and corrosive aspects of social media and now we've got this incredible strain of AI to contend with that will make 'fake news' on social media look like children's stories. For two decades now companies like Google and Facebook have been given a free pass to gather our personal data, run insane psychological tests on their user-base without seeking any kind of permission (one of the reasons Tristan Harris left his job with Google) and now just because Microsoft got there first, the Goog have got their digital knickers in a twist and are scrambling to unleash this incredibly dangerous technology onto the masses. There has been zero oversight of the big FAANG tech companies since 2000 and now these LLMs are being unleashed on us with zero oversight there too. This cannot end well.
And there will continue to be lax oversight because now we're continuing to compete in a global economy where China, Russia, Saudi Arabia, and other heavily state controlled networks have competing LLMs. It's the cold war all over again, now with more propaganda right under your nose. Anonymity is on the chopping block.
Yep. It is a bit of a shit show, gotta love human ego and greed.
A shitshow on top of a shitshow. Imma go eat an entire pint of Ben and Jerry's RN.
Would be easy to spiral and say ‘omg I agree.’ Let’s step back and look at the big picture. Human history has been a shitshow… like we have not made the best decisions. Cult leader worship since caveman days, a guy nailed to a tree, KONY 2012. Does humanity forget history too?
The question posed at the beginning is designed to scaremonger I guess those in power. But the question put forward was not ‘Will AI create human mass-extinction?’ It was: "What probability do you put on human inability to control future advanced Al systems causing human extinction or similarly permanent and severe disempowerment of the human species?"
Al Impacts Survey of 738 ML researchers from June - August 2022
Is anyone else getting a sinking sensation? This presentation clearly explores mind-blowing insights with irrevocably far reaching (perhaps civilizational) consequences yet has only 39k views. The muted response is more frightening than the video.
I have that sensation since text-to-image models got released. Everyone was shouting "We will automate the boringness!" and then they automated one of the most creative endeavours there is... I spent some time in the comment sections under artist's videos, and young students of art are having existential crisis over this. Nobody stopped to think, "Do we really need this ?"
@@Hexanitrobenzene man, this is so serious and so not discussed. $#@!
Be quite peasant. The ai has been here for 30 years, the infrastructure just wasn't built for it to move around....... go read!!!!!!
Within seconds of the video he starts making misleading statements. 😅
"50% of AI Researchers think there is a 10% chance of blah blah blah"
Ah... so of 100% _of all AI researchers_ likely think it is about a 5% chance, +- some deviation percentage.
Ya, if you feel a sinking feeling, you aren't betting on the overall thoughts of 100% of researchers.. because as a group, they only see a 5% chance of it going bad.
But hey... lots of spots on the lecture and corp tour circuit open to talk about AI.
Ambulance chasers, fear pays.
Oh I'm a subscriber, this is hilarious.
@@thethree60fiveput another way, the median prediction was 10%. Not at all misleading. I'd argue you're being misleading, that math you're doing is not at all how stats works.
Also a reminder that this is the most extreme possible outcome. This totally ignores the lesser but still bad outcomes. Like....the entire content of this video
Thank you for sharing this thought-provoking discussion on the risks and challenges posed by existing AI capabilities. It's crucial to address these concerns and ensure the deployment of AI is accompanied by adequate safety measures. Upgrading our institutions to navigate a post-AI world is an important consideration for shaping the future of large-language model AIs responsibly.
Oh boy you did not understand.
Over the course of this video, my mind was definitely changed. Incredible.
It’s hard to even partially process this for me. More astonishing is that there are not sirens going off everywhere. I had no idea most of this was happening the way it is. Why is this not the worlds #1 priority? Where are our leaders? Feel kinda sick right now…
money
To be clear, China has taken steps to protect itself from AI. Everyone else? Guess we're just rolling the dice. And no the laws in China aren't really good but at least they're aware of the danger. Meanwhile America just points and laughs at our leader because he's old.
Our leaders are exactly where we put them doing what they've always done. People need to stop voting for the establishment and learn to recognize the establishment. They are extremely easy to spot when you understand how they are two sides of the same coin.
men have been geared towards the detriment of others since the garden of eden. in reality we have been living in a fantasy as millennials in the west. Good times don't last long looking back at history. The good news is Jesus really did rise as a matter of history, and he offers a way out of this mess
lol as if world leaders actually care about solving anything if there is no money behind it.
More people need to watch this video, such an amazing presentation. Thank you for taking the time and effort to talk to us about this important topic.
I have watched 2 of his video back to back and CZcams recommended me this one even before search for it. I was about to search this video as it was mention in previous video
Please index this video for easy search and access to the diverse topics tackled in the conference.
As a software engineer, I'm EXCLUSIVELY interested in AI safety!!! I'm planning to pivot my career into AI for this very reason because tragically it's not going away and there are no global EMPs scheduled for our planet.
Same here. I'm finishing all that I'm currently doing and pivoting to AI safety in the near future
Hello , my little sister wants t be a software engineer, With the progression of AI will software engineers still be needed?
@@khanyanyameni9578 This is an excellent question. I actually asked an AI this very question a couple weeks ago, to which it replied that more engineers will be needed in the future to maintain these AI systems.
While nobody can be certain, I tend to agree that this will indeed be the case, at least for several years to come. The real trouble comes when robotics catches up and gives bodies to these AI. While that may take some years, should AI be combined with advances in Quantum Computing, and it almost certainly will be, these AIs will almost certainly be able to design and build their own bodies, and if AGI is actually possible, well... you better seek God real quick if that happens. As for me, my faith and trust is in Jesus Christ and his salvific work on the Holy Cross, so I literally fear nothing that man can devise.
That said, AI is truly remarkable, even as it stands now with these advanced LLMs can do. It should give us great pause and we must constantly remind ourselves that we are indeed having a dialog with a machine processing language through a series of algorithms trained to recognize complex semantics based on a model created from trillions of English sentences across millions of topics.
@@khanyanyameni9578 Everyone is in denial about this, I think. Companies have already started slashing way back on labor and wages in favor of dirt cheap AI tools. Developers and engineers are still around at the moment, but that's going to cut way, way back in the future. US companies already started hiring cheap software engineers and developers overseas after firing all their domestic ones.
The safest bet right now would be to divert to information security, network engineering, or cyber security. Those are going to be areas that AI won't ever be able to truly replace people. Everyone else dealing with machine languages though is in for a rude wakeup call I think. The days of high salary pay for those jobs is ending soon.
@@graceoverall I see AI designing synthetic cells that will merge with human bio cells and build a biosynthetic human...Homo AI Cyborgis...the next transition for humans. I see the humans of the future without the need for smartphones tablets or computers. I see humans of the future being connected to an AI central node. We become the smartphone computer etc. In the future.
Thanks to everyone at CHT for this video.
excellent presentation
real
I agree 👏
just two comment let me fix that
We got to see this masterpiece becuz u shared. 🙏
ey
Finally someone explained how CZcams can read my mind! It knows what I have looked at and what I've been thinking about. Thats so amazing. AI is so intelligent!
This needs more views, the more people that see it the better
We all need to watch and understand this presentation thank you for making it free for all of us.
Thank you for raising the alarms. You are level-headed, provide factual evidence to support your claims, and calmly explain one of the scariest introductions of technology in human history. It's hard not to become hysterical at some of the possibilities. Just like the gospel of Christ, we can only tell and share the truth, it is up to the rest of humanity to come to terms with the truth. Some will and some won't but we have done our duty.
The insane thing is, i was negatively looking at the "8 months ago" date of this video before watching it today in December 2023.
Because AI is moving SO FAST now, that i was already thinking "oh so they're not gonna talk about GPT4 or Gemini".
Isnt that incredible, and scary?
Saw this in January 24😵 is it outdated??? Self compassion was his very important word at the end of the speech. One of the best speeches, I ever heard.
38:28
"Give a man a fish and you feed him for a day;
Teach a man to fish and you feed him for a lifetime;
Teach an AI to fish, and it will teach itself biology, chemistry, oceanography, evolutionary theory… and fish all the fish to extinction."
Incredible and much needed discussion. Thank you
😷
but HOW? AI doesn't control an army of advanced autonomous robots, its BS.
@@koffing2073 I also thought the parable was ridiculous... but kinda expresses the continuous self improvement (why would you fish the fish to extinction? Surely not as stupid as humans.) these models can go through.
@@stephentrueman4843 Yes AI would also manage the livestock to assure a constant supply, until we can finally print food that is as good as the real thing without animal suffering.
This is just a perspective, AI could also choose to preserve fish better due to understanding it more.
it's mind-blowing that this only has 38k views. maybe it's time to get back in touch with Netflix, guys?
IMHO, A great presentation overall. One thing that stood out for me (among several things) is @ 32:50. What I heard is that AI has learned how to "think outside of the box" (Develops Theory of Mind). Yes, my comment is very generalized, but I hope you get the overall sense of it. It's also one of the scariest comments I heard in the entire video, although some other comments/observations came very close!!
This was very deep and well Said. I just can't believe it's already over a year old!
I don't usually go in for that 'take a breath' stuff. But, in this instance....I genuinely stopped and took that breath with you. This is entirely conceptually overwhelming, and to attempt to take a step forward without risking humanity falling off the cliff...the margin of risk is almost invisible until we've all but hit the pavement.
heyy you might like "Simple Mindfulness Strategy -- Take 5 " also
We hit it… we shouldn’t have started. Even the idea of Maki g robots with human faces is bizarre and not okay.
@@unclejesse4271 lol that is so benign, funny comment
😂
I, too, took a breath each time the presenters suggested it, and each time my thought was, "How far is AGI advancing in the time it takes me to take this breath?"
It is unbelievable how calm guys are and still managed to show the urgency of the issue.
Incredible! Also loved the massage.
What I hear is: first people must get awaken (enlighten/kinder, smarter, get rid of ego) then AI will not be used against each other. It is a great tool to improve our lives.
If we have to get rid of ego, then all is lost.
@barry hernandez I was thinking the same thing. Humans have been talking about that stuff since before recorded time.. there's no chance we don't screw ourselves with this technology. I mean we botched even just the technology we already have. Everything is weaponized and monetized. Now I get it, "first world problems" but fr with the technology we have, we should be supporting life even more and even better than ever before. And yet, here we are, famined and impoverished, uneducated and divided.. real food and Healthcare is gate kept from the majority.. but yeah, let's make programs that learn from the great examples we set. "Do as I say, not as I do". If we can't even make that work in closing the "generational gap", how are we going to make it work with an intelligence that lacks intuition, nuance or compassion?
@@nintendbro3275
Yep. You get it. The big picture. As you said we pretty much weponize everything available, what makes anyone think this time will be different. However it's to late now. It's just going to have to play out.
@@Cleatus546 1. not possible, 2. hahahahahaha - even if ...
@@Cleatus546 BTW: I did not buy that computer can "recover "original image by reading the brain scan MRI.... seems that something is missing there... quite a lot actually...
I have a feeling that absolutely nothing will be done and in 10 or 20 years we'll look back at this video like "I cant believe they were right all this time and we did nothing"
We may not be around.
We will look back and hope we could still be here while hiding in a cave without electricity or food avoiding drones strikes.
same.
The ship is sinking and the passengers are dancing.
And if we are predicting 10 or 20 years, according to this video and our inability yo forecast exponential upon exponential, that could be 2-3 years… or maybe even sooner
Good job guys, definitely eye opening didn’t even think about an arms race between companies to implement AI into products.
To be fair here, we've already been using this for about a decade (things like SIRI, automobile computers, GPUs, etc.). For whatever reasons, the use of "AI" exploded in marketing over the last year, like suddenly a memo went out to all the companies to use the new buzzword "AI."
But on the other side of that same coin, the programs did explode in ability over this last 6 months. It's like they reached a tipping point and flew off the rails. We absolutely should not be using it in our daily lives yet, if ever. The negatives far outweigh the positives, especially when it comes to things like labor and human capital. Companies have been salivating at the ability to fire everyone they can and use cheap "AI/ML" tools as replacements.
Thank you for speaking on this
I lost track of the amount of times my mind was blown just in the first half-hour. Excellent presentation!
Extremely well articulated discussion of the possible dangers of Ai. Highly recommend people watch this.
Please make more vids like this! I love your artistic + analytical approach. Tried the monochrome look the other day and felt so fancy. Thanks for the idea!
The potential in my own field of expertise (psychology) is significant. When I think of how this technology could reduce human suffering my mind races with the possibility. This technology will inevitably mean a paradigm shift across just about every field of human knowledge, not just psychology....and not just science. The implications of that are more than I can process right now quite frankly but I will share the video because I'm not sure what else to do.
very likely you won't be needed at all like as many of us, money dictates and people might even find it easier to disclose their troubles to a non human entity, time will tell, but for sure we face something serious
CZcams recommendations needs to foreground this. We have a year at most until the moment has past and the enmeshment they describe has happened. I am forwarding this to everyone I know.
I've been thinking about the possible harms of AI but this presentation just made it real for me. It feels weird because this only happens in movies but, this is the most important moment in our human history...we need to get it right. If we don't, we're done.
Maybe AI will protect us from AI
If we can't control the first AI, how will we control the second AI, which presumably will need to be more powerful in order to control the first one?
@@KaLaka16 I agree. Fight fire with fire. I hope we figure it out.
@@goodlookinouthomie1757 It may not be so much about control as it is about alliance. Anything may emerge as the models get larger it seems.
We’re done. Why are people acting like it’s a mystery what happens next.
What I would like from these educational talks is more info about the processes/mechanisms for regulation. I do know there are smart people talking about this but I think it needs to be part of these types of Education events
It stretches every boundary to beyond a point most people understand. The leap forward in A.I. in say the last 7 years is astonishing.
This was the best collection of current applications and future risks of AI so far. Info about AI safety on this level needs to be part of the general conversation society has asap.
This is actually the best presentation ever made so far.
nah (but it was great)
I bet chat GPT can do a better one... 😞
It really is. They actually get the tech and how it might change the world. Other people are just seeing better code and content. They are looking several steps ahead.
Thank you for sharing.
thank you