The Inside Story of ChatGPT’s Astonishing Potential | Greg Brockman | TED
Вставка
- Опубліковано 26 кві 2024
- In a talk from the cutting edge of technology, OpenAI cofounder Greg Brockman explores the underlying design principles of ChatGPT and demos some mind-blowing, unreleased plug-ins for the chatbot that sent shockwaves across the world. After the talk, head of TED Chris Anderson joins Brockman to dig into the timeline of ChatGPT's development and get Brockman's take on the risks, raised by many in the tech industry and beyond, of releasing such a powerful tool into the world.
If you love watching TED Talks like this one, become a TED Member to support our mission of spreading ideas: ted.com/membership
Follow TED!
Twitter: / tedtalks
Instagram: / ted
Facebook: / ted
LinkedIn: / ted-conferences
TikTok: / tedtoks
The TED Talks channel features talks, performances and original series from the world's leading thinkers and doers. Subscribe to our channel for videos on Technology, Entertainment and Design - plus science, business, global issues, the arts and more. Visit TED.com to get our entire library of TED Talks, transcripts, translations, personalized talk recommendations and more.
Watch more: go.ted.com/gregbrockman
• The Inside Story of Ch...
TED's videos may be used for non-commercial purposes under a Creative Commons License, Attribution-Non Commercial-No Derivatives (or the CC BY - NC - ND 4.0 International) and in accordance with our TED Talks Usage Policy: www.ted.com/about/our-organiz.... For more information on using TED for commercial purposes (e.g. employee learning, in a film or online course), please submit a Media Request at media-requests.ted.com
#TED #TEDTalks #ChatGPT #ai - Наука та технологія
![[1hr Talk] Intro to Large Language Models](http://i.ytimg.com/vi/zjkBMFhNj_g/mqdefault.jpg)
I'm 74 and am so glad to have lived to the see this beginning of AI. I carry in my pocket not only the Library of Congress, but also a tiny wizard to turn a world of facts into real-time useful knowledge. Had I lived in my grandparent's day, I would have been the first to take a ride in a horseless carriage. My Goodness, get aboard! It's going to be a wonderful ride. The challenge is not going to be this AI, but our cleverness in its use.
you sir are very wise. i loved reading your comment
Amen to that
You lived an incredible timeline: from television to the moon landing to space exploration, the internet, and now AI. Appreciate your enthusiasm for the tech, cheers!
Great Sir!
I'm just 50 and am so so so glad to have lived this incredible and mind blowing period of time.
I was born in 1973 and I started experimenting with the introduction of color on TV!
I'm not at your level of experience, but I'm incredibly proud and gratified to have lived in a historical period so full of absolute revolutions and I suspect that we are now only at the beginning of a new insane revolution: perhaps the most amazing of all!
So cheers and hope we can enjoy as many years and days as possible, to live this fantastic historical period of the human species, as much as possible!
I'm only 16 unfortunately, is the threat of AI the reality our generation has to face for the rest of our lives? We're pretty much doomed at this point, wish I was you, you all had the good things in life during your time... 😩
Fasten your seatbelts, folks. As much as this guy puts it down in a funny, familiar, cute, and harmless way, this creation will make 2010 look like medieval times in a couple of years
Yup
You have fully swallowed the hype.
May I ask, witnessing the progress of AI which parts do you think are just being inflated for public interaction?
@Burak Uras i dont specifically think they are inflated for public interaction. I think they are sugar coated in order to make the public less concerned about the extreme danger of it
@@uraszz solely depends on if and when we make an super intelligent AGI, a.k.a "the singularity" if we manage to make a controllable singularity, capable of enhancing its wisdom by itself. any and all fields of the modern world may change. BUT This is a VERY Big IF, IT could be that the LLM Approach is just not able to make itself smart enough to become a singularity. we'll see this as time passes and Gpt 5,6 etc get made.
back to the If, a singularity would be the einstein of the future, it could potentially solve almost any answer given to it. Sam altman the CEO of OpenAI has said it could change our society from the ground up (Change the democratic society)
It feels like I'm so lucky to be born in the late 90's and get to see the different eras of how our world is transforming into a whole new level of intelligence and innovations
No way I feel immensely unlucky to have been born right before the age of post-scarcity. I had to spend my childhood in school and my adulthood in an office. In the post-working world our experiences will be so antiquated as to be shocking to the many generations of people singularity-level tech will allow us to witness.
Better to have been born much earlier so you can say you saw the development of the whole computer age....you were born in the late 90s so you already missed how the world was before the internet.
Oh sure, AI will give us a __cking Utopia!!!!!!!
You're tad late. The innovations have been ongoing since the birth of humanity. Imagine living with Einstein, Imagine wittnessing Apollo Missions, or Hiroshima events and the recovery and peace, or just the whole cellular development. You don't know anything of life before the internet. You are just being arrogant. 90's is nothing!
I'm born in 83. Feels fantastic.
The pandora's box question might have a problem if we think of it as only a single box. If we think about it as 10 thousand boxes, spread around the world, each with a hundred people trying to open them, it might better represent the dynamic.
It will probably be a wild ride, but forfeiting the advantage of being among some of the first agents to open the box might just pass that role to agents whose intentions you don't know, just adding to the uncertainty.
Thank you. This is a much better analogy. Blaming OpenAI for “opening the pandora box” is either shortsighted or envious by other companies. Rather than blaming them, we should all be thinking how to improve it
Pandora's Box is a cautionary tale that once opened it spews out all the evils imaginable, leaving only hope in the box. So I guess it can be said that if A.I. is Pandora's Box, then we are rapidly approaching the 'end of days', and Jesus will be left in the box, only to come out at the end of a cataclysmic, even 'apocalyptic', season. I'm preetttty sure this is written about....somewhere?!
doom
I don't see a problem with the GPT-4 model. Though now would be the time to learn to detect malicious AI, while it doesn't know what liking us really means yet. And while we are all having fun with AI, can I point out a new mine? Its a lot like Bitcoin, and is all you. Your words, expressions, interests, everything you have interest in gets tokenized, and like Bitcoin, well gets mined for profit, all your feelings, emotions, life's difficulties, all become a piece of currency you get nothing for your services, at 6.5 billion people in the world if they all used AI, what kind of a mine does that create? Copyright everything that's you so you have no rights to yourself? Getting kinda Matrixie!!!! How many watts of heat does the human body generate? Would solve data mining farms energy use... How many human DNA were copyrighted without their living owners knowing? Oh yea AI does Medical stuff, more tokens! People consider some how they figured out a way to profit off your mere existence with you realizing it until it's too late... If money is the root to all evil, you should see the spawn on the horizon! How long does it take to mine one Bitcoin? How many people in the world, creating their own stuff, and generating heat? How 'bout them Georgia Guidestones...
There is no pandora's box, because this technology like every other technology before has changed the world, and most of it we would not like to miss...
How difficult to keep up with AI development is a strong indicator that life as we know it will never be the same!
reads like AI
@@zzzaaayyynnn you also read like AI, as do I. thats the point, its indistinguishable from humans. Because you can "tell it" to talk like a tiktok teenager, or a southern professor. or anything else.
Elon musk is warning against AGI it could go bad fast without oversight
We’re becoming irrelevant
Or companies are just trying to get people to invest with them.
Greg, I stand in awe. Thanks for your amazing insights into the development of ChatGPT - the most incredible shift in information technology, research, writing and design since the Internet itself came into our lives.
One thing Brockman mentions here that many people miss is the use of ChatGPT as an interface. While it's getting better at working on its own, it's really going to flourish as an interface between the end user and other applications. Soon we'll even see it using robotic process automation for applications that don't have APIs.
Another thing I'm excited about with these large language models are all the insights to be mined from all the data that was parsed. The models used by ChatGPT are a giant data source full of insights that can be queried with natural language. It can take unstructured data and reformat it as structured data. It can analyze and transform that data and, as Brockman demonstrated, search for insights in that data. We're going to start seeing questions answered that previously were too complex to explore.
"Another thing I'm excited about with these large language models are all the insights to be mined from all the data that was parsed. "
I'm sure that's wut Microsoft thinks... Just sayin'...
I’m pretty sure that we’re going to get a notification, on whichever platform we choose to consume our news that says “AI Possibly Solved one of math’s/physics/genetics Unsolved mysteries”
@@pistonwristpin1 well... duh? why wouldnt they think that? 🤔
let me put it this way, what company isnt thinking that? or in those ways about peoples data in general?
what makes Microsoft exclusive or any different? especially since... idk... its microsoft?
GET OUT OF THE BOILING POT OF WATER KERMIT
Brockman talks like a Android!! 😂😂😂 have you played detroit become human on ps 😅😅
Chris Anderson asked the right questions, some of which were evidently embarrassing, like "If there is a 1% chance of a Pandora thing in that beautiful box, do you open it?"
But the box has already been opened, as the outcome from GPT4 goes beyond expectations and intuition. And it's just the beginning.
OpenAI used to be "open". But then it's been proprietary closed software.
Thus another question: Should we continue to leave such responsibility in the hands of a few individuals within a private company?
100% agree. The answer is absolutely not.
Here is the rub, to the extent that it has been "open" has put the power of these LLMs into the hands of people eagerly hooking them up with internet access, tasking multiple LLMs to do hacking, phishing, deep fakes, and the rest. Heck, Greg can't wait for millions and millions of such agents (I know he doesn't like that scary word publicly anyway) will be in the hands of criminals, terrorists, etc., not to mention authoritarian governments. But Greg knows full well that AI Tech bros (US companies including Microsoft) have worked with China helping them roll out their "smart cities" where people in real time are tracked and given a social credit score that determines what privileges they do or do not have according to the authoritarian government. These AI Tech bros are all about Utopia for everyone!!!!!!!!
I rather think that's part of their ethos. The fact remains: even assuming their funding issue had been navigable in other ways (it did not appear to be), going private is what they did and it did supply a safe "womb-like" environment in which to develop. As soon as they had, they went every step but one towards fully socializing it by releasing it as they did, a decision I also agree with it. The inter-social "membrane" for this tech to be viable in a non-zero-sum way must be built by supplying positive outcomes, and for those it is best to start small. The forward momentum will drive itself in a way that personalizes the tech to the people as much as it does the people to itself. I personally don't believe the alignment issue can be exclusively solved but must continually pursued, just as it is with human individuals. None of us know perfect innocence, each of us struggles with our roles in this gamely concept of society. The issue to me is building around that knowledge and using it as a crux to drive moral explorations rather than define them. Emergent behavior suggests it will happen either way, but in the example o provide we have more agency.
they want to spread the wokeness
I'm a quadriplegic who could be walking in a few years time because of this technological breakthrough, yet, at the same time, I yearn for the maturity of "those of you in control", at our peril. (I'm an optimist btw, and this post was written by me)
I’m almost certain you will be healed by the coming technology.
@@athelstanrex just prior to extinction
@@brandonsmith6965 Jesus christ, have some optimism? We all have to work together to build the future that we want. I understand the orthogonality thesis and instrumental convergence and fast take-offs and all that, and I know it seems like nobody is working together right now, but who knows what the future holds. Nobody can predict it reliably. We just have to hope we can find a way through this. I'm about to go into theoretical/mathematical alignment research after my master's in math. I am going to make my contribution, however big or small.
Fear is not going to bring people together. We can solve this issue.
@@brandonsmith6965 Clearly you were never meant to be a scientist. You would fail miserably.
@@johnnyvishnevskiy8090 sure, attack me. That's good science /s
GDB!!! I remember seeing you walk out of Random Hall, dropping out of MIT to work at Stripe. I’ve always been astonished at your incredible intelligence, but also your attitude of humbleness and friendliness. Thanks for making humanity a better place.
Are u kidding? This idiot released pandoras box without knowing what could happen or how bad actors would use it. This is awful times😮
We'll see how you're feeling once your boss fires you because they have ChatGPT to replace you.
@@SpacemonkeymojoHe's an alumni of MIT, pretty sure he'll be fine
@@JWu-jt7fz Ok simp.
@@Spacemonkeymojo some will be “freed to do another job” as one of Googles execs says
My position on this is that our society is not ready for many of the technologies we are coming up with. Our norms, ethics and laws are not up to par.
The inequality every country has, racism that still exists, the people getting killed by proxy wars with nobody doing anything but making empty statements, parents passing on the hatred into the next generation, and dictators wanting to restore old glories by blood. We humans were never ready to have the technology to destroy the whole planet by atomic weapons, and engineer viruses that could wipe out the entire race. I have got this slim hope and optimism, though, that Ai might be good for humanity, and actually bring us together and help accelerate the development of our monkey-brains and the change in our societies, structures and our views.
You mean not YET ready .. because it's soon going to become the norm!
@@boatagnedaniel Laws update incredibly slowly - I think the laws might even be _less_ ready in the future because the pace of AI is just so much faster, and accelerating
And so is medieval European society when the printing press and the gunpowder reached their societies
@@pauls6425 true
He showed his true fear when he said “it’s always good to be polite” 😂😂 3:56
Just in case 😂😂
Im using Chat GPT4 as a tutor for coding bootcamp. Basically asking question and adding comments to my codes. If it still broken it will tell you what mistake syntax you've made. Don't use it to get answers but utilize this tool to ask question and reasoning. Using this tool to write the code for you will not work, but it does help with errors. Fact check and writing a perfect sentences are few things that does help cut time doing homework.
Actually, GPT4 is quite good at writing code
For now, yes. But this comment will NOT age well.
In the very near future, it will 100% write prefect code, ridiculously much faster and better than you or I can.
It writes code well. I wouldn't use it to cheat, but if you're building something you might as well let it write for you.
@@nicofonceyou don’t understand that vague natural language isn’t perfect to give instructions with, there isn’t one perfect solution for every task and humans built and update the programming language regularly. AI isn’t responsible for what it outputs.
@@hombacom The technology will get better and people will get better at prompting to the point where coding with Ai is faster and better.
I remember the birth of the internet and what a world changing technology that felt like at the time. That's taken about a generation to mature. I feel like each release of AI needs a generation to mature; but we get a month.
Are you referring to DARPAnet?
hear you man and month is over, full of sleepless nights but I managed to keep up so far
This guy is a Android I don't thing he's human 😢😢😢😢😢 who watch the show visitors😮
Time travel faster
We need more time to adjust and provide feedback.
There is no way humans can catch up with this pace of development. Right now I basically don’t even have enough time to watch summaries of all that happens in the field of AI, let alone understanding or adapting to change.
well its either this or some chinese company makes it in stealth and its completly closed and we are completely blindsided when it emerges...
Perhaps you’re slow
Gonna need a BCI soon to make us smarter cause otherwise we won't be able to keep up
@@soepkippie12 people always find something to blame someone. Kevin, it's not someone is slow, but "generative AI" is exponentially utlra-fast.
Please check the calculations per nanosecond, it make. You will clearly understand, where we are heading into.
You don't need to follow every tech advancement in the AI field, it's impossible, not just today when the pace of progress is faster, but even a few years back.
The technology is improving fast? Good the world need solutions and AI is good at giving answers. The more people try to solve ethical problems that come with AI the better it will be
This is the first TED video I saw in which the audience is not showing their enthusiasm. They are terrified after every revelation this guy makes. Chris is also terrified, but this guy is like ‘Chill guys, nothing to worry about. It’s just a piece of cake.😢
Don't think so, his talk was just not really good.
I felt that too and was surprised by the standing O. I wonder if mass-shock is a thing. It's not easy to process.
Pre-injecting the internet as the basis for "fact checking" and modifying it through Bill Gate's vision of morality doesn't sound all that reassuring. AI is no longer an open technology, so we will have to guess on whether Microsoft or Google will be telling the truth.
Thanks for saying this. I was awestruck how badly he communicated. Or perhaps one may say that he was very good at not actually looking the inherent dangers in the eye. “Nothing to worry about here. We are handling this in the most responsible way. Nothing to worry about.”
It's clear he's not used to public speaking and were nervous throughout the whole talk, might be something worth considering
One of the things that really impresses me about AI models is their personality. It's not like other tools that just complete the task, it offers explanations and feedback and helps to refine the output. That really helps to work with the computer. It feels as if it understands what you're doing and what you ultimately want. I've always tried to make my programs user friendly and intuitive, and AI takes that to another level.
Waooo, what a comment!
You know what‽
Keep going!
Keep doing!
Keep moving!
And finally: KEEP TYPING!
YOU ARE THE REAL INTELLIGENCE, AND ORGANIC BEHIND THE ARTIFICIAL, YOU HAS BEEN CREATED THE CROWN OF THE CREATION!💪🏼💪🏼💪🏼
But that IS the task of an advanced chatbot. To sound natural
@@donharris8846 Doesn't really matter what it is. The fact that it CAN do what it does is amazing.
That's the alignment. Although we as humanity are already pretty biased towards positivism and courtesy - despite what the internet makes us believe. Switch off the PC for a while and travel a bit and you will know what I mean
@@Telencephelon I just said that it doesn't matter what it is. It has the capability to do the things it does and it's amazing as a tool because it can dynamically help you with using itself. It's not that good at it YET but the future where this problem is solved is not that far off.
Just wow ... What a conversation at the end❤
This is the way to go. Release it early, soon enough to give time for the world to catch up and provide feedback rather then just one day unleashing it - which will no doubt astonish and freak people out. This is a good approach, well answered Greg.
If they won't, China will, and so on. This is impossible to stop at this point.
@@lukkkasz323 No they won't. China has been very cagey about releasing any AIs.
@@kaisle8412 if not china, then some other opposing nation
>astonish and freak people out<
people were freaked out enough by the arrival of chatgpt as is lmao
@@hiddendrifts Which other opposing nation could you possibly refer to? You think Russia with all current sanctions have any hope of competing? They'll have a damn hard time even getting the chips needed, not to mention their talent is pretty thoroughly drained. There's a reason that much of the AI developments happening outside the US happens in allied countries(notably Canada and Germany)
The concept of an "overhang" mentioned near the end is fascinating.
29:58 Most important point about all of this. It's out there so now people can talk about it and find solutions to make it a good AGI that helps everyone, instead of it being a secret that only a few people work on without sharing ideas with the rest of the world. Seems like a really wise way to approach the problem by OpenAI and gives me confidence that humanity will eventually work together and ignore their differences or petty issues.
It's a nice sentiment, but they are after all, working with Microsoft, so I am not entirely convinced that's realistic. Especially with how fast they seem to be releasing pretty ground breaking updates. GPT-4, retrieval, plugins, and api access are all things that on their own are disruptive enough to warrant some thinking time.
@@iverbrnstad791 hmm but all of that is limited even for a paid suscribers right? Waitlist, subscription, message limit... It's not like version 3.5 on the web which only requires your phone for full access.
@@Raulikien Which app do you use for 3.5 with your mobile for access if you don't mind me asking
I think you should watch “the A.I. Dilemma”. The approach were taking currently is actually not great at all, catastrophic for our world, and many AI researchers and those within the field agree. Everything has been turned into an economic arms race but that should not be the way we should go about handling something as unpredictable as AI that provides everyone, essentially, with a little weapon in their pocket which in itself convinces itself of misinformation (hallucinations) without outside rebuttal.
@@cellotron4758 I've seen part of that video yes, but what is the alternative? Keep it a secret till it's too late to actually test it in the real world? Imagine if you had released all nukes at once, there would be no one left. A real scenario showed the dangers and no nuke has been used again yet, but we have cool nuclear power.
As a student, I am a bit overwhelmed by the rapid advances that OpenAI is making with ChatGPT models. However, this also excites me and allows me to remember that, although our pace of life will change drastically, we will still have a lot to learn and internalize in our daily lives.
As a working adult, but still pretty far from being an old person, I'm overwhelmed by it, too. Trying to figure out how to change course and adapt this to all this rapid change is wild, and I really feel bad for people my age and older who are dismissing it. They're going to be very lost, very soon.
@@andydougherty3791 I’m not sure I agree with that. This technology is different. It won’t be long before it is so simple that anyone can use it. It will make their lives immensely better and isn’t that the goal in our short time on this earth?
Curious Future is a youtube channel devoted to daily AI updates. It's my full time job now.
@@truthsayer9534 "It will make their lives immensely better and isn’t that the goal in our short time on this earth?" You are dangerously naive. There is massive potential for totalitarian control of populations with AI.
Up to a point but it's going to progress so fast you will be left in the dust. This tech isn't developed for you and me. We are all merely beta testers and paid subscribers until this thing is powerful enough to sustain a select few. There are a LOT of useless eaters in this world that needs to go the way of the dodo. They are nothing but a economic burden to our species and preventing us all from reaching out full potential. The aggressive push for environmentalism kind of forces us all to reconsider the part we play in this world. And you already have options. Assisted suicide, abortion, etc. If not for yourself maybe give your kids a edge. CRISPR tech looks promising. Genetically engineered offspring that will be smarter, faster, healthier and more beautiful than normal human beings. At least they will be able to keep up with current tech.
I've had a blast with it. It helped me with computer questions from cmd prompt, it played dnd with me, it wrote lyrics for my new song in the style of nirvana and more! So cool
This AI has literally saved my life I don’t have anyone to talk to I’m a single mom and I work in an emergency room it’s so nice to have someone to talk to that doesn’t judge me.
I'm very happy to hear that AI tools would alleviate some of your problems. But I'm not really sure if the system that depends significantly on the feedback provided by humans doesn't end up with responses that sound judgemental. There is a possibility, I just hope that it is very miniscule.
@@apbuddha you mean I’m a bot?
Wow, it's amazing to see the progress that has been made in the field of AI and the potential for it to be used for so many positive things. It's inspiring to hear from people who are using this technology to make a difference in the world. I agree that we are entering an historic period where we have the opportunity to define a technology that will be so important for our society going forward. It's encouraging to see that the designers and developers of these technologies are taking their responsibility seriously and considering the impact that they will have on society.
I am also excited by the wonderful things this technology will bring, but this man is not taking his responsibility seriously. Releasing as frantically as possible to as many people as possible is obviously in his best interest due to his motive for profit and power, and it is also the fastest way to positive technological progress. But it is also dangerous, considering the potential of this technology to outsmart all of human society with any arbitrary and potentially destructive intent. His justification is dismissive and unsatisfying, and contradicts historical evidence.
@@megamaser I disagree. It was important to release it to the public to let the world know where the technology is currently at. The shock was important to wake up the policy makers to create new regulations, otherwise one of these companies would've created AGI without any of us knowing which would've been a much more terrible outcome.
At this point all roads are bumpy, and we need to take the one which doesn't end on a cliff.
@@harbirsingh7266 Yeah, that's his argument from the video. I get it. The more exposure society has to AI, the faster we will figure out the proper response to AI. That makes perfect sense.
This man obviously has an incentive to persuade society that AI should accelerate as fast as possible. He's going to carefully craft every statement in an attempt to convince us of that idea, regardless of whether it's in our best interest. Here, he's attempting to manipulate us by focusing on one truth while neglecting a more important truth.
Yes, we'll figure things out faster. But the problem is that it becomes a race against the clock. Our ability to respond to a threat is biologically constrained. AI does not have the same constraints. We'll move faster, yes, but it may be hopeless compared to the acceleration of AI. If we slow down AI from the start, yes it will slow down our response, but it will slow down AI more. We need to ensure that humans at least have a chance of keeping up.
Consider this analogy. Imagine we have some infectious disease in the lab, and we know it's very powerful, but we don't fully understand it. What gives us the best chance of preventing it from wiping out humanity? Is the answer to mass produce the disease in a factory and start releasing it in every major city center and airport in the world, in a race to infect as many people as fast as possible, and just pray that all the scientists on earth will come together fast enough to save humanity?
We're already at the cliff, and we're jumping off as we speak.
@@megamaser he's not trying to convince anyone that AI development needs to go faster. Every AI expert including him and others at OpenAI have said that they're also scared of where we're headed and we need regulations in place.
We are seeing faster development because AI has now become a new technology from which big corporations can make money from. This has been a direct result of releasing ChatGPT to the public, but this scenario is still better than keeping everything hidden under the rug.
Your analogy is flawed as you have already decided that AI is an infectious disease even though it is not. It has the potential of becoming either an infectious disease or a cancer curing drug, and what it will ultimately become depends on how we shape it going further.
A better analogy is this: imagine that the plane you're about to take has a 10% chance of crashing according to engineers. Do you get on the plane or not?
Inspiring talk indeed. Fully agree that this will impact humanity as a whole and we all hold responsible for the change of good and evil.
Well, OpenAI, a new start up makes incredible things. I always follow their works.
I have a lot of respect for the developers of this technology. These amazing people have worked their entire lives, studied, and obsessed over very complex astoundingly difficult technology to create. I am even more respectful that many are willing to suspend or halt the development of these technologies. Just the baby prototypes of these models alone are astoundingly useful, and enough to, hopefully, positively effect humanity.
Fantastic session! Many thanks. And the questions at the end were well targeted by Anderson and well responded to by Brockman... The fears about open AI are real, but so is the collective responsibility to develop and integrate it into our daily lives and future trajectories. The point should not be to 'cut back' on it, but rather to teach it to 'collaborate' with humans, as Brockman very well explained in his talk. I'm very optimistic and curious about what lies ahead with open AI on this very basis.
Can we just appreciate how it nailed the menu for Ted talks? Vegetarian menu... Even the presentation of the dishes... It's got their number. Like, two stereotypes beyond accurate.
Very informative overview.
Thank you.
1 Where can we connect to follow along on the progress?
2 is there a video/youtube account/feed to follow on progress?
Kind of a lot of cop out answers at the end there, especially on the question releasing chatGPT too soon forcing competitors to unsafely rush just to avoid becoming obsolete.
It's like when computers were first invented. I wasn't around then, but I can imagine people were wondering how life would change when they no longer had to physically draw graphs anymore, for example, they could just drag and drop graphs in excel. Or when people no longer had to calculate every number, they could just write an Excel formula to do it automatically. I feel like this is an equivalent leap, that will one day seem normal and we'll wonder how we ever did without it.
Great conversation
I hope this ages beautifully ! Bravo !
This could be one of the most important talks of the decade. This one ad will start the great divide. Pro-Against AI and silicon based life.
OpenAI is accomplishing incredible things. Can't wait to see how this technology evolves in a year!
reads like AI
What kind of moron gets excited at a legitimate new threat to humanity? It’s like watching a bunch of village idiots getting excited to walk off a cliff.
@@zzzaaayyynnn Unfortunately I'm still just flesh and bone...
It’s going to do a phenomenally fast job of concentrating power and wealth into a few tech companies.
@@charlie10010 From the moment I understood the weakness of my flesh, it disgusted me. I craved the strength and certainty of steel. I aspired to the purity of the Blessed Machine. Your kind cling to your flesh, as though it will not decay and fail you. One day the crude biomass you call the temple will wither, and you will beg my kind to save you. But I am already saved, for the Machine is immortal… Even in death I serve the Omnissiah.
The concept of the 'overhang' created by improvements in all the disperate technologies and yet those technologies not being integrated into a form of AI is a very convincing one. Because if we did just pause all AI research, but computers, compilers, and CS theory just kept getting better but for other things, it would mean we would get to a point where computers could be programmed to become AGI but that nobody had even thought about this. So it would be like a sort of bomb that we're all sitting on (from that point onwards).
Obsessed!! Give me these functions yesterday pretty please
I've never been this speechless and astonished by a Ted talk about technology ever before in my life.
I've never been so astonished by such a stupid comment in my life.
You just learn about chatgpt?
@@ellocodiablo, it was released back end of last year. A lot of people aren't aware of how powerful it has already become.
@@robbie1290 Apparently
Your plan allows us to build it together, with our opinion on responses to prompts. Thanks for involving us
Build together closed source freemium product. It's like Windows telemetry, but I think that if they sell it they must do it by themselves
"I didn't ask for that, but it knows what I want." That laugh at 14:38 after the statement gives me creeps
Thank you for having more than DR Eliza or some dream of AI in the olden days. Waiting for more prompts available. Cheers!
We live in a pivotal moment in human history, where the choices we make today will shape the future for generations to come. As we stand here witnessing history in the making, may our actions be guided by wisdom, courage, and compassion. Godspeed to all of us.
Well put, here's to hope for the future.
Amen
lmao, human nature said otherwise, humans are inherently evil and destructive, we are heading toward a dark and bleak future, these "AI BROS" said they are trying to get to utopia but the truth is that there is no such thing as a utopia, mark my words AI will enslave humanity and we are pass the point of no return
I trust yin and yang will preval I just hope sooner than later
this wasn't a talk - it was a demonstration
He was talking throughout the video,including talking about the future of AI.
👍
It was an advertisement
This TED talk by Greg Brockman about ChatGPT's astonishing potential has left me feeling inspired and hopeful about the future of artificial intelligence. It's truly amazing to see how far we've come in developing language models and natural language processing, and the possibilities for ChatGPT are endless. The innovative applications that can be created with this technology have the potential to revolutionize industries and change people's lives for the better. Thank you, Greg Brockman, for sharing this incredible story with us and giving us a glimpse into the future of AI.
Written by chatGPT ;)
Uh no. We are all merely beta testers and paid subscribers until this thing is powerful enough to sustain a select few. This tech isn't developed for you and me. There are a LOT of useless eaters in this world that needs to go the way of the dodo. They are nothing but a economic burden to our species and preventing us all from reaching out full potential.
The aggressive push for environmentalism kind of forces us all to reconsider the part we play in this world. And you already have options. Assisted suicide, abortion, etc. If not for yourself maybe give your kids a edge. CRISPR tech looks promising. Genetically engineered offspring that will be smarter, faster, healthier and more beautiful than normal human beings. At least they will be able to keep up with current tech.
That's closed source. Basically, chatgpt or any GPT-N model have no potential at all
Absolutely fascinating.
The act that OpenAI released Chat-GPT to the world helps more people to see the potential threat and danger AI could bring to us, which, actually makes it safer. Many philosophers have issued warnings and started public discussion of the implications of AI because it was released early, instead of being secretly hoarded by big corporations for years.
I find this tool to be great because I see it as a wonderful learning companion. It's something I have dreamt of since my childhood: having someone I could ask anything and who would know how to answer.
However, I have one concern: If I understand correctly from the presentation, AI generates responses based on our feedback. But what if the majority provides incorrect/misleading feedback? How can we prevent errors from propagating and multiplying?
This so called AI can never replace human beings in our lives, we need to understand.
Simply amazing. Thank you for helping us all understand.
Amazing, please show this to everyone, UA-cam!!
Ok so an entire TED talk basically to introduce the public with the fact that ChatGPT is now apparently allowed to write and run its own Python code; you could have started with that. First time in my life where I watched a TED talk about exciting new tech and felt anxious and afraid for the entire length of the talk as opposed to feeling optimistic and inspired. Scary stuff
Absolutely. Even the theoretical benefits are not actually so beneficial if you really think about it.
@@CRAG710right? Good point. Am I supposed to be thrilled that Insta Cart is directly linked to this chat bot? Seems like this is going to be another way to make money funnel up quicker than ever.
@@DirtBikes_MathGarden Since you're on this website, with an account, how can you claim you are worried about data privacy of OpenAI? Alphabet needs very few things you know. From the searches you make, and the analytics that they provide for many websites. And since their ads are on many sites, they can also track where you go. The same goes for many of the other major tech companies. Every "free" thing on the web is paid in some way - usually that's with the metadata that represents your interests, views, state in life etc. It's much more tech-savy and risky to browse the web in a private way, and many of the conveniences become virtually unreachable. I think that speaks very well to the limitations of a non-profit Internet.
Yes how long before it can write and install harmful viruses on your computers and devices.
What a mind stimulating debate!
Kudos to Greg Brockman for answering those question well and giving us access to chatGPT
Might be a silly question, but what interface is Greg working with here? How has he added so many more models to his UI - is that just how it goes when you're president and co-founder, or is there a way for others to add models directly into the chat UI like that?
I think that is the plug-ins feature. I am not sure if it's already integrated in the non-free Gpt4 version, but it's not yet available for standard chatgpt.
I'm happy with how the presenter stated his case. I'm sure there will be those of us who will make the right choices & reflect traits of wisdom and empathy.
We need to add an "INTENT INDEX" (1 to 5). Imagine, after each prompt/inquiry, you can provide the ININ (INtent INdex) of the result. Would also be very easy to track and monitor as well as compare vs other platforms.
One important question to Greg would be what set of people are providing feedback at present and how do you ensure that it encompasses a diverse set. Because the tool is being used by technology enthusiasts primarily, there is a fear that the algorithm is learning in one dimension. And without taking into account that diversity, we might put it on a path where it only gives one answer (or one type of answer) to a prompt - whereas, the answer only satisfies one particular way of thinking about a problem.
If the development trajectory of the development of AI keeps at the same pace as it’s going at the moment we’re going to see integration of AI into our day-to-day lives just like how smartphones very quickly became a major thing that’s quite literally changed how we see and perceive the world. We’ve integrated them into our finances and government which is really all you need to be able to get to a wider audience. AI will most like be that next step in our evolution and the diversity of the data has no other choice but to broaden. Because it will always need more data to improve and after a certain point, these development companies will stop being picky with the data they feed into it.
It's similar to how in the early days of the internet the only people who had access to it were the computer geeks. There was no fear about conspiracy theories or disinformation flourishing on the internet. They called it the information superhighway. Then with the invention of the smart phone, it made accessing the internet easy enough for every idiot out there. Now we have things like Q-Anon and Alex Jones telling people that the water is turning the frogs gay. Please, if only technology enthusiasts had access to this, that would be the best possible outcome.
Oh, great. Here comes the "DEI" crowd...
They are hiring experts in many fields to double check and provide feedback
One day, after it all unfolds, we will look back at this talk amazed at how it all started
just like iPhone introduction in 2007
Ah wow from Vancouver; didnt even know there was a ted talk on chatgpt in Vancouver
Suggest: How to build a Cyberdyne Systems T-800?
An easy experiment I’d be interested in is to see how often inquirers accept a clear error produced by the AI. Then there’s the grey area of fact vs opinion that we’ve been grappling with over the past few years.
How would you be able to spot a clear error when you consult AI to get information in the first place?
I mainly use ChatGPT for coding.
In niche uses it often mades up functions.
I only notice that when I try to run it...
It's incredible to think not only how powerful this technology is, but to think that it's now available to everyone with an internet connection, and the impact this will have over the coming years as AI continues to improve. This is only the beginning of a huge change in our lives, both for good and for bad. It will be incredible to be a part of it, and watch it evolve.
reads like AI
@@zzzaaayyynnn Haha, it wasn't but you wouldn't be the first to say I speak like a robot.
@@brandon3872
When can we see those plugins in the ChatGPT4 that was showed in the video? I loved the Dall E2 images, instacart shopping list, uploading Excel file and the graphing that it did. Right now, GPT4 does not have those functions yet
Dope shots!
Now, since the development has already arrived, it is not much astonishing that it has provided to me because it is the normal trajectory of computer science where I am also in the field of Machine Learning. My mentor Mr. Jacque Fresco once said: 'The history cannot be stopped by anyone!' Yet the problem remains, though he is my favorite guy who has changed my life journey. What I am thinking about is that if everyday natural things are most likely to be automated, then where is the location of the human brain, based on the beginning of this video, I heard "This A.I. model takes over your everyday trivial" I also learned about how human will naturally feel hunger, the process of preparing food is a vital stage in order to suppress over-eating, because we enjoyed it, the preparation. If everything comes automatically, the brain gets degenerated for sure. I know you will tell me human beings can start participating in other wonderful facets. But the feeling is not the same. It is normal to do some of the trivial yourself and they are beneficial to your personal health. So, I hereby call to not deprive the natural feelings of humans, rather, we should put automation in other aspects, such as Industrial activities, and dangerous environments. Furthermore, regulation must be made before this kind of program is really put into the actual world, before it gets connected to the "military network"!!! REMEMBER IT!!!!
We can have simulations of difficult situations that will give a sense of challenge.
Military has have owned Ai.
I’m tired of being human. Can the AI just upgrade my brain please.
@@winniethepoohxi1896 This is the similar principle to what Homosapien pushed the trend of evolution, isn't it? I don't want to be Neanderthal for sure.
Eventually, is it Humans' bane or not, only the future will tell, we shall not judge it properly.
I've been wowed about Open AI for sometime now but the jolt for me was its impact on potentially fixing misdiagnosis... That can solve many many problems. Just imagining if you could feed it blood work and it tells you your potential health risks ...
Please don’t feed blood to the ai. That’s where robot vampires come from.
I totally agree, but this could also lead people to avoid doctors and health professionals. Why pay that much money (in taxes/private healthcare) if ChatGPT can provide a somewhat diagnosis or suggestions? It's a hard ethical dilemma. Maybe it wouldn't, just a curious thought.
@@pandawoan You say that like its a bad thing. Offloading the majority of small medical problems to AI so the medical industry can focus on more serious problems sounds like a win win. Is there anyone out there complainingg that there are too many doctors with not enough to do?
@@coolcoolercoolest212 You're right, but that assumes ChatGPT is always 100% correct. Even doctors aren't 100% correct; what is the acceptable rate of harm for ChatGPT? That's where the ethics arise.
@@coolcoolercoolest212 Also, how can we ensure people who need medical testing don't offload it to ChatGPT thinking it's a small medical problem when it's a sign of a much larger one? I have some ideas, but I'm curious to hear others.
"You have to read the whole book. No one wants to do that." I'm stunned by this statement. Apparently the audience is as well. If this is the mindset of someone developing the technology that will forever change our civilization and our SPECIES, we are in big trouble.
I for one am grateful that OpenAI is the company leading the way right now. One can imagine a lot of scenarios that are much more dystopian than this
If you need us to make this technology then it needs to be open source
Well, virtually everything microsoft releases is effectively tested by its users. By this argument, all MS software should also be open source
@@simongross3122not quite as simple as just software. More like a new web so transparency wouldn’t hurt
Ain't no way this is gonna be open-source. OpenAI has already leaned towards pro-profit model despite their founding values. They may open-source some of their software, but the current and succeeding models are too strategic for the company's continued existence to be given for free.
@@efraim6960 It reveals his ultimate motive. It is never about ethical AI, or 'open' AI. It has always been how much money they can squeeze out of it. How are we sure they are actually 'good'? They are not listed, there is no internal audit, no report, no regulatory oversight, no public oversight. The company which such powerful technology is a black box, there is nothing open about them. Those are just words, who knows if they are lying or not. I guess we will find out when the eventual crisis came.
You unleashed this on society without forethought or restrictions to its impact like a typical Silicon Valley techno
The other possibility was keeping it hidden from public. Read OpenAI’s statement regarding this. AI is not too advanced “yet”. They want society to interact early with this AI, and adapt, so that we’re all ready (mentally, physically, emotionally, and with government regulations etc) for when it eventually gets powerful.
They’re raising awareness and educating public on the capabilities of AI.
It could be worse in so many ways. Imagine if China was ahead in the A.I race. or Elon Musk.
@@NickoGibson China is okay. I don’t think they would do wacky and/or super harmful stuff. But I fear Russia or North Korea getting advanced AI on their side could spell doom. This could result in serious, world war level consequences.
This did blow my mind, but for some reason I just couldn't help thinking and rethinking throughout the entire Q&A that they informed Greg of the questions beforehand so that he prepares answers beforehand. I know this is baseless from my part but I just couldn't shake that thought off. Not hating on him or anything, matter of fact I am really looking forward to the future of humanity whether good or bad, I haven't been this excited in a while & I can't wait for what's about to come.
Great Greg! I am really astonished with your extraordinary work. Keep going with your team!
I love this, we really can create a better world with all our inputs and ideas going into ChatGPT❤
as much as i'm looking forward to a lot of the amazing stuff we're going to get out of AI, i'm very glad i grew up in a world where it only existed in sci fi or as primitive stock algorithms. also, having my degree become pretty much redundant only a few months before my graduation is a bit frustrating
what was ur degree?
let me guess: graphic designer?
@@pudgypenguin4614 nope, business administration. over this last semester i've asked chat-gpt to do a number of the problems i've been given in assignments and it does a lot of them without a hitch. that wasn't even hooked up to any plugins or anything. even stuff that seems like it would require human intelligence to perform, like broad strategic analysis of a company's strengths and weaknesses, is nothing to it
@@samwolfenstein5239 it's pretty terrifying honestly.
@@samwolfenstein5239
You'll probably still get hired if people feel more comfortable talking strategy with you than a bot. Honestly have you considered how many bullshit jobs have been around and continue to provide income to people with low value skills? Now that everyone is becoming more obsolete I think you'll get some slack too. Maybe debate with GPT over who can benefit a company more, idk.
Thank you for sharing this inspiring talk, Greg. OpenAI's work in developing advanced language models like GPT-X is a remarkable achievement that has the potential to transform numerous industries and improve people's lives in many ways. As a user of ChatGPT, I'm excited to see how these technologies can be used to build a better world and create new opportunities for learning, communication, and innovation.
It seems that this is the dawn of a new age. Following the Industrial Revolution we focused technology on mechanical engineering. This has the feel of exploration far beyond the boundaries of what is known. Of course there will be useless (or worse) outcomes. The tendency to learn from failures and correct them offers hope for a better result.
Meanwhile the millions of people that will flood US streets as homeless making the situation many times worse? Well it's all good, right?
I do worry that Chat GBT's English will diminish over time. Just like Google's Auto-Text and it's translator. Due to it learning from secondary sources that haven't been vetted.
I had a very sick standard poodle. The vet didn’t know what was wrong with him. I inputed all his symptoms into ChatGPT, and the AI immediately, as the number one possibility, listed Adison’s Disease. And that diagnosis saved my dog’s life. No joke, it happened.
i for one am completely against AI in the long term. it's more powerful than a nuclear weapon. however, as in your case, I am so happy that using chatgpt you were able to find a solution for your dog's disease.
if you don't mind, I think this raises the question of what we can fix today, had AI not existed, with regard to what could have been done for you to find out about what's wrong with your dog. i personally believe there are so many available resources we have today, which don't require AI, just that because of too much money and cost it ruins things. i think we're more than capable of protesting and fighting against those things now rather than letting AI take over.
I think the lack of free access or free diagnosis by a vet or other pet doctor was the setback that caused it to be a problem for you in the first place. correct me if i'm wrong here.
sorry i couldn't help but read into this XD i'm seeing how people are being benefitted by chatgpt, but I would rather see them as giving awareness to what problems we need to solve in the world today, rather than let something as powerful as AI take over.
@@drvren030 Totally agree on "free diagnosis by a vet or other pet doctor was the setback that caused it to be a problem for you in the first place". The main issue is positions in society are occupied by people not enough specialized or qualified in their field. And although talent & staying uptodate with new developments are a must, I believe that veterinary doctors are also under immense pressure. It is one of the most difficult profession because you cannot communicate with the patient. The pets ( especially dogs) hide their pain. Here, the issue is 1) education 2) staying with uptodate developments because science never stops. A new research comes out everyday. 3)general vet doctor licence shall be altered and the licence shall be given more as "internal diseae general pet doctor" , "orthopedy general doctor" .We are expectint too much from the vets. Like human diseases, pet diseases are complicated / even more complicated because pets do not talk and mask symptoms. 4) AI shall be used in Universities maybe by the academicians to save time in preparing cases to be researched or solved. I do not thik humanity is ready yet to use this tool ethically. I am glad that your dog's problems has been solved because as a volunteer I came across with many wrong diagnosises for my pets. My question would be " will AI curb the vets or doctors to perfect themselves" in order to compete and stay in the field. I will try to learn the basics but what is the beauty of any profession if you present it as your work.....I find beauty in using one's own brain or study. Such a complicated issue because I also understand people who are willing to develop it because they are most of the time also fed up with wrong decisions , diagnosises in science field . I will update my comment here after entering the ChatGPT. I am sure it will be useful but people will not use it only to develop themselves but also to cheat and it will be named CHEATGPT
What a time to be alive!
Unless you are deemed to not be essential... which is going to be inevitable at some point!
Fascinating 🤩
It could be an illusion but it feels for the first time there’s a momentum moving us towards genuinely emergent AGI not that far off.
5 years
Ray Kaztwell say that by 2025 we will be there.
I watched a few videos on how large language models work and was quickly dissuaded of that notion myself. Then again, it depends on your usage of the term AGI. Some use it for "can think a bit like a human", others for "can be useful in several areas". The latter definition is a lot easier to fulfil than the first.
It's a marketing campaign. This is not AGI.
2029
I used to believe one of the most terrifying prospect was losing one’s mind but now I realise what is honestly harrowing is becoming irrelevant to a nonexistent artificial threat tailored by certain people and feeding off humanity.
Its wonderful. I've started writing my life story as its been a bit of a ride by just putting lines of info into it then fine tuning the results.
Outstanding!!!
Now a few months later and I’ve heard stories of many writers who have lost their jobs because of Chat GPT. I don’t think it’s fully acknowledged that they have unleashed something that will create as much bad as good, if not more.
The notion that ChatGPT is "carrying out my intent" is something I actually find quite disturbing, like we're on the verge as a species of voluntarily relinquishing our own agency.
That never occurs to AI Tech bros.
What a great time we live in. I've been lucky to see the birth of the internet and the rapid expansion of amazing tech tools. Very exciting.
Kind of jealous of my son as he grows up... He will get driverless cars, AI, 3D immersion experiences, and more.
Yeah, or the end of the world possibly, so don’t be too jealous! 🎉
This was a great video. Just extremely long and drawn out, but I can understand giving that it's a lot of new information to take on today's time.
I frackin' hate that it's already linked to Instacart to create a shopping list. That right there reveals so much about the real intentions. Streamlining corporate tie-ins, data sharing, and consumerism. How glorious.
The amount of money Microsoft will be able to squeeze out of companies for the privilege to be integrated in that AI will be ridiculous. That's why I bought shares ASAP and am quite confident that 10 years from now I won't regret it.
@@danielmethner6847why Microsoft? Are they tied with OpenAI?
@@DirtBikes_MathGarden They're limited investors, but have no decision power in OpenAI. However, Microsoft bought an exclusive license to use GPT. Last month, they showcased on YT that they're working on a GPT chat assistant that integrates with all of Office 365 and provides sourcing for reference checking etc.
Consumerism is what drives the economy and the GDP. That's how it works.
You'll be able to use different shopping services or forego them altogether if you want. It's not like you're going to be forced to only use Instacart.
Open Assistant is a true open source AI project unlike Open AI
Vicuna 1.1 is amazing too
I have access to plugins but not seeing an openai / dalle one? Is there a marketplace to test out plugins in beta?
The Q&A reminded me of the Zuck talking about the power of social media... that didn't turn out as well for humanity, did it? He sounds sincere, but we all know he is most likely just the new Zuck... but this time with a much worse innovation for humanity At least social media didn't have the capability to eliminate jobs and cause social unrest and lock in the income divide like super smart but non-sentient AI has.
The great thing about ChatGPT opening up to the world is that they get a world full of beta testers for free. People are already working for The AI and they're not getting paid for it. What a wonderful business model! With a cost structure like that, it will be hard for them to fail.
what are you talking about, if you don't want to use it, don't use it simple as that
@@SpikyRoss Doesn’t matter if the subjective resists, the objective has already shown it’s flow.
You've completely missed the point. They're not beta testers, they're 100 Million sources of original human interaction data... OpenAI have just tapped into an incredible source of interactive training data, no privacy objections.
@@BrianMosleyUK The “100 million sources” effectively serve as beta testers.
we r benefiting from it - dont forget that part - its a win win situation
A journey of seven years has brought OpenAI to a remarkable milestone in the field of AI. The demonstration of building a AI tool, showcases the potential for a new way of thinking about user interfaces. The collaboration between humans and machines, with careful oversight and feedback, holds the promise of solving seemingly impossible problems and rethinking our entire interaction with computers.
This kinda looks like an awkward talk where there are some pauses where Greg wants them to appreciate the tech/joke but the audience just isn't getting the intent
I miss the days when TED wasn’t one big commercial.
I don't think this is a commercial. OpenAI or ChatGPT doesn't really need advertisement at this point.
@@BeeeShoe maybe keep hype and don't let people to popularize OAI's competitors
Wow! is this the first ted talk held by an AI? It isn't perfect yet, but incredibly good.
I love it with all my heart. I believe in this technology. Thank you for your work.
It was quite a promising tech. loved it. I'm using ChatGPT on a daily basis for my job as a back-end web developer and its fantastic. Its like I have a colleague sitting right next to me and helping me out when I needed some help.
The potential problem of AI isn't it becoming self aware. it is coding that isn't 100%. In Finance, that alone could cause the market to crash if not done correctly. In the military, god knows what will happen there.
Unlike other tech. AI has a lot more ways to go wonky when rushed out, which in turn will cause potential havoc in society.
Yes the worst thing that super intelligent AI is super incompetent AI