Sam Altman and Bill Gates Talk AI | Is GPT-5 close?

Summary

В этом захватывающем видео Билл Гейтс беседует с Сэмом Альтманом о будущем искусственного интеллекта и его влиянии на общество. Обсуждается, как технологии могут изменить наше понимание цели человечества и организации общества в условиях изобилия. Альтман делится своим видением о том, как AI может радикально улучшить продуктивность в программировании, здравоохранении и образовании, а также о возвращении к проектам в области робототехники. Вдобавок, они касаются темы конкуренции в разработке AI и важности международного регулирования мощных вычислительных кластеров. Завершается диалог размышлениями о потенциальных сообщениях, которые человечество могло бы отправить своему создателю через достижения в квантовой механике.

Takeaways

• 😀 Сэм Алтман говорит, что ограничения GPT-4 скорее всего будут устранены в GPT-5
• 😊 Билл Гейтс и Сэм Алтман обсуждают будущее ИИ и как ИИ повлияет на мир
• 🤔 Сэм Алтман советует стартапам оптимизировать контекст, а не поведение
• 😮 API Открытого ИИ будет быстрее, надежнее и дешевле, но всегда будет компромисс между производительностью и стоимостью
• 😀 Сэм Алтман женился и сказал ИИ "да" на церемонии
• 🤔 Стартапам не стоит фокусироваться на устранении ограничений GPT-4, они будут решены в GPT-5
• 😮 Кто-то нашел слабость ЧатГПТ - если показать ему определенное изображение, он перестает работать
• 🤔 Андрей Карпати объясняет смысл жизни через взаимодействие с создателем
• 😊 Билл Гейтс считает, что ИИ принесет огромную пользу, несмотря на трудности
• 🤔 Сэм Алтман говорит, стартапам стоит создавать продукты с учетом скорого появления ИИ уровня человека

Outlines

00:00

🤖 Размышления о будущем AI и человеческом предназначении

В этом разделе обсуждается влияние искусственного интеллекта (AI) на будущее человечества, выделяя философские и психологические вызовы, связанные с прогрессом технологий. В частности, рассматривается перспектива, в которой AI может взять на себя задачи, традиционно выполняемые людьми, оставляя человеку больше времени для занятий, не требующих интеллектуального усилия. Это порождает вопросы о человеческом предназначении и организации общества в условиях, когда не существует дефицита учителей, врачей и хороших идей. Автор поднимает проблему адаптации общества к эре пост-дефицита и выражает уверенность в необходимости принятия технологических изменений, несмотря на предстоящие трудности.

05:00

🚀 Импульс для стартапов в эру GPT-5 и AGI

В этом разделе делается акцент на значении искусственного общего интеллекта (AGI) и следующих версий GPT для стартапов, с особым вниманием к ожиданиям от GPT-5. Сэм Олтман на мероприятии для стартапов подчеркивает необходимость адаптации к быстро развивающимся технологиям, предлагая строить компании с учетом скорого достижения AGI и улучшений в GPT-5 по сравнению с текущими ограничениями GPT-4. Основные советы касаются оптимизации контекста вместо поведения, баланса между производительностью и стоимостью в технологиях и фокуса на разработке продуктов, не ограничивающихся текущими ограничениями GPT-4. Также упоминается «криптонит» для GPT, вызывающий сбои в работе модели, и личные достижения Сэма Олтмана, включая его свадьбу.

Highlights

The video is abnormal, and we are working hard to fix it.
Please replace the link and try again.

Transcripts

there's the chance of the system taking

control and for some reasons I'm I'm

less concerned about that I'm glad other

people are the one that that sort of

befuddles me is human purpose when the

machine says to me Bill go play pickle

ball I've got malar radication you're

just a slow thinker then you know it is

a philosophically confusing thing and

how you organized Society yes we're

going to improve education but education

to do what there's a lot of

psychologically difficult parts of

working on the technology but this is

the for me the most difficult cuz I also

yeah satisfaction from that and it's

like in some real sense this might be

like the last hard thing I ever do well

our minds are so organized around

scarcity scarcity of teachers and

doctors and good ideas I do wonder if a

generation that grows up without that

scarcity will find the philosophical

notion of how to organize society and

what to do maybe they'll come up

solution and I I'm afraid my mind is so

shaped around scarcity I I even to have

a hard time thinking of it if we can get

into this world of post scarcity we will

find new things to do they'll feel very

different but I think the only way out

is through we just have to go do this

thing it's going to happen this is like

now an Unstoppable technological course

the value is too great and I'm pretty

confident very confident we'll make it

work but it does feel like it's going to

all be so different yeah I think if it

comes across as asking for a Slowdown

that'll be really hard if it instead

says okay do what you want but any

compute cluster above a certain

extremely high power threshold and given

the cost here we're talking maybe five

in the world something like that any

cluster like that has got to submit to

the equivalent of international weapons

inspectors and the model there has to be

made available for safety audit pass

some tests during training and before

deployment that feels possible to me I

wasn't that sure before but I did a big

trip around the world this year talked

to heads of state in many of the

countries that would need to participate

in this and there was like almost

Universal support for it right now I

guess the we're looking a lot of

productivity improvement from AI which

you know that's overwhelmingly a very

good thing which areas are you most

excited about yeah so first of all I

always think it's worth remembering that

we're just sort of on this long

continuous curve um so like right now we

have an we have ai systems that can do

tasks they certainly can't do jobs but

they can do tasks and there's

productivity gain there eventually

they'll be able to do more things that

we think of like a job today um and

we'll of course find new jobs and better

jobs and I totally believe that if you

give people way more powerful tools it's

not just they can work a little faster

they can do qualitatively different

things and so you know right now maybe

we can speed up a programmer 3x it's

about what we see I mean that's one of

the categories that we're most excited

about it's it working super well but if

you make a programmer three times more

effective it's not just that they can

write they can do three times more stuff

it's that they can at that higher level

of abstraction using more of their brain

power they can now think of totally

different things and it's like you know

going from Punch Cards higher level

languages didn't just let us program a

little faster and let us do these

qualitatively new things but I think

it's worth always putting in context of

this technology that at least for the

next five or 10 years will be on very

steep Improvement curve um these are the

stupidest the models will ever be but

coding is probably the area the single

area from a productivity gain we're most

excited about today Healthcare and

education are two things that are coming

up that curve that we're very excited

about too so how do you see robotics

super excited for that we started robots

too early um and so we had to put that

project on on hold it was hard for the

wrong reasons it wasn't helping us make

progress with the difficult parts of the

ml research and you know we were like

dealing with bad simulators and breaking

tendons and things like that and also we

realized more and more over time that

what we really first needed was

Intelligence and cognition and then we

could figure out how to adapt it to

physicality and it was easier to start

with that with the way we've built these

language models but we have always

planned to come back to it we've started

investing a little bit in robotics

companies I think on the physical

Hardware side there's finally for the

first time that I've ever seen really

exciting new platforms being built there

and at some point we will be able to use

our models as you were saying with their

language understanding and future video

understanding to say all right like

let's do amazing things with a with a

robot and what about the competition is

that kind of a fun thing that you know

many people are working on this all at

once it's you know both like annoying

and motivating and fun I'm sure you felt

similarly but it does push us to be

better and do faster and we're very

confident in our approach we have like a

lot of people that I think are like

skating to where the puck was and we're

going to where the puck is going and

feels all right Bill Gates interviews

Sam Alman on Bill's podcast called

unconfuse me which was apparently on

YouTube this whole time it's pretty

interesting it's pretty General they

don't go too deep into anything but it

is an interesting thing to watch those

two kind of discuss where they think AGI

AI where the world is going and how AI

will affected but one thing that really

caught my attention didn't happen at

that interview but it sounds like it was

happening while that interview was being

released so this is kickoff for YC Y

combinator where they get a lot of the

founders of these startups it sounds

like it took place at opening ice

headquarters in San Francisco Sam Alman

was there now for the life of me I can't

find any video of this I can barely find

any images but in his sort of initial

speech to the new Founders to these Tech

Founders that are all building their own

companies samman suggested people build

with the mindset that gpc5 and AGI will

be achieved relatively soon that most

GPT 4 limitations will get fixed in GPT

5 here are some key takeaways one

startups need contextual optimization

more than Behavior optimization

providing more information via rag Etc

can be more beneficial than fine-tuning

two openai API will continue to be

faster more reliable and cheaper however

there will always be a balance between

performance and cost for example iPhone

will keep its one one and a half day

battery life to optimize for performance

even though Battery Technology has

improved significantly and three it's

not advisable to build companies that

Focus primarily on addressing current

GPT 4 limitations most limitations will

get partially SL entirely fixed in GPT 5

this was published in the standard a few

days earlier Chad GPT 5 release date

what we know about Open the Eyes next

chatbot in recent months hype has been

building around a new and more powerful

version of the tech that underpins Chad

GPT the language model that will be

dubbed at GPT 5 Altman allegedly spoke

about gbt 5 and gbt 6 during a talk at

his former Venture Capital firm Y

combinator at the alumni reunion in

September according to two people who

attended the event he said that both the

AI models were in the bag there's not

too many more details about one it's

going to be available so we'll see are

we going to see GPT 4.5 are we going to

see GPT 5 we have yet to see in other

news there's this apparently someone

found the Kryptonite to chbt apparently

if you upload this this image to Chad

GPT and ask it what do you see or

something to that effect it just breaks

why who knows a lot of people reporting

being able to break Chad gbt with that

with that image I don't know why in

other news Sam Alman got married but yes

it's true Sam Alman did say AI do like I

do a I do I don't know it's really late

right now either way congratulations to

him that's it for me for today here's

Andre karpathy explaining the meaning of

life in it is really interesting to

think about like what the puzzle of the

universe is did the creator of the

universe uh give us a message like for

example in the book contact UM Carl Sean

uh there's a message for human for any

civilization in the digits in the

expansion of Pi in base 11 eventually

which is kind of interesting thought uh

maybe maybe we're supposed to be giving

a message to our creator maybe we're

supposed to somehow create some kind of

a quantum mechanical system that alerts

them to our intelligent presence here

because if you think about it from their

perspective it's just s quantum field

Theory massive like silure tonaton like

thing and like how do you even notice

that we exist you might not even be able

to pick us up in that simulation and so

how do you uh how do you prove that you

exist that you're intelligent and that

you're part of the universe