Sam Altman talks with MIT President about AI (May 2024)

Summary

MITの18代目の総長であるサリー・コーエンと、OpenAIの共同創設者でありCEOであるサム・アルトマンが対話し、AIの将来性と社会的影響について議論しました。コーエンはAIが社会に広く利益をもたらすべくMITが取り組んでいると証言し、アルトマンはAIの進化とそれに伴う職業消失、新しい職業の創出について語りました。二人はAIによる教育の変革、科学的発見の促進、個人のプライバシーとAIのバランスについても洞察を共有しました。この会話は、AI技術の進歩がもたらす可能性と課題に興味を持つ聴衆にとって非常に興味深いものとなっています。

Takeaways

• 😀 サリー・コーエンは2023年1月にMITの18代目の総長に就任し、その卓越した管理者として知られています。
• 🧠 サム・アルトマンはオープンAIの共同創設者であり、CEOでもあります。彼らのミッションは人工知能を通じて人類全員に利益をもたらすことです。
• 🤖 AIの「絶滅の可能性」という問いに対して、サムは0ではないと答えますが、現在はまだ非常に初期の段階だと考えています。
• 🔍 サムはAIの進化について、過去10年間でより現実的になり、技術革新の一つとして捉えられるようになったと語っています。
• 🛠️ AIのバイアスを取り除くためには、システムを特定の価値観に従わせる必要があるとサムは考えています。
• 🏥 AIは医療など特定の分野で、人間のバイアスを超える可能性があるとサムは述べています。
• 🔒 AIの進化と個人情報のプライバシーに関するバランスが重要な課題であるとサムは指摘しています。
• 🌐 オープンAIは完全にオープンソースではないが、他の方法で公開し、広く利用可能にしています。
• 💡 AIの発展は科学技術の進歩に重要な役割を果たし、持続可能な経済成長を促進するでしょう。
• 🚀 AIの進歩は人類の生活と仕事に大きな影響を与えるが、人間は創造性と有用性を追求し続けるでしょう。

Q & A

サリー・コーエン氏はMITの第18代総長として、どのような背景を持っていますか？

サリー・コーエン氏は、セル生物学者であり、デューク大学で8年間副学長を務め、優れた管理者、創造的な問題解決者、教員の優れたさと学生の福祉を推進するリーディングアドボケートとして知られています。MITでの多くのイニシアチブの中で、人工知能（AI）にも焦点を当てており、AIが社会全体に幅広く有益になるようにMITの努力を進めています。

サム・アルトマン氏はどのようにAIの可能性を考えていますか？

サム・アルトマン氏は、AIが社会に積極的な影響を与えるために非常に重要なツールであると考えており、AIの発展を通じて人類の技術的な進化を促進したいと考えています。また、AIが持つ可能性を最大限に活かすために、安全性とユーザビリティのバランスを慎重に考えていく必要があると述べています。

AIの進化によって生じる偏りやバイアスを取り除くために何が必要ですか？

AIの進化に伴い、システムを特定の価値観に沿って行動するように調整する必要があります。しかし、誰がバイアスの意味や価値観を決定すべきか、そしてシステムがどのようにすべきかを決定するために、社会的な合意が必要です。また、AIシステムは人間に比べてバイアスを少なく保有できる可能性があり、そのように努力することが重要です。

AIの発展と個人のプライバシーの関係についてどのように考えていますか？

AIの発展は個人のプライバシーと密接な関係がありますが、プライバシーとユーティリティ、そして安全性の間のトレードオフを適切にナビゲートする必要があります。将来的には、AIが個人の生活全体にアクセスできるようなサービスが登場する可能性があり、それに伴いプライバシーに関する新たな定義や取り決めが必要になるでしょう。

オープンAIはどのようにして他の人々にAIの恩恵を提供していますか？

オープンAIは、無料で利用できる優れたAIツールを提供し、何百万人、将来的には何億人もの人々が使用できるようにしています。広告を出さず、公的な善として提供し、人々が手軽に利用できるように努めています。また、AIを通じて人々が生活をよりよく、より豊かにすることを目指しています。

AIの発展によって生じるエネルギー消費の増加と、気候変動との戦いへの貢献という間の緊張感はどのように解消されますか？

AIは非常に大きなエネルギーを必要としますが、AIを用いて非炭素基盤のエネルギーや、より良い炭素捕捉技術を開発することができれば、そのエネルギー消費は大きな勝利になるでしょう。また、AIは遠隔地での仕事を可能にし、エネルギーを節約する助けにもなります。

AIが科学や工学分野にどのように貢献する予定ですか？

オープンAIは、AIを用いて科学発見の速度を増やし、人類の進歩のコアエンジンを支えることに興味を持っています。また、教育やビジネス、コンシューマ分野にも貢献する可能性がありますが、科学分野でのAIの影響は特に期待されています。

若手研究者や起業家に対してアドバイスをいただけますか？

現在はキャリアをスタートさせるのに非常に興奮的な時代であり、AIに関連する何かを行うことを強くお勧めします。また、自分のキャリアで何が最も興味深いかを早期に把握し、それに向かって動くことが大切です。また、テクノロジーの豊かな将来を目指して、人類全体の生活を向上させることに貢献することが重要です。

AIが金融分野にどのような影響を与えると予想されますか？

AIは他のすべての分野と同じくらい金融分野にも大きな影響を与えると予想されますが、具体的な変化については詳細は述べられていません。教育やヘルスケアのような特定の分野でのAIの活用については、すでに多くの可能性が見込まれています。

AIの発展による偏りやバイアスの排除とプライバシーの保護はどのようにバランスを取るべきですか？

AIの発展による偏りやバイアスの排除には、システムを特定の価値観に従わせる必要がありますが、その決定は社会的な合意に基づいて行われる必要があります。プライバシーの保護については、ユーティリティと安全性の間のトレードオフを適切に管理し、個人のプライバシーを保護しながらもAIの恩恵を享受できるようにすることが求められます。

オープンAIは今後どのような方向发展を予定していますか？

オープンAIは、AIの能力を大幅に向上させ続け、それを誰もが手軽に利用できるようにすることを目指しています。AGI（人工一般知能）という言葉は使いづらく、代わりに特定の能力X、Y、Zを達成するまでの時間を考える方が良いとされています。今後もAIの恩恵を広く提供し、経済的な価値を創造するシステムを目指して開発を進めていく予定です。

Outlines

00:00

😀 トップクラスの対話

MITの18代目の総長である生物学者のサリー・コーエンと、OpenAIの共同創設者でCEOであるサンジャイ・アルマンが対話に臨む。コーエンは2023年にMITに就任し、多くのイニシアチブに取り組む一方で、AIの社会的貢献にも力を入れている。一方、アルマンは技術革新を通じて人類全般に貢献することを目指す。対話は、多くの学生からの質問を通じて行われる。

05:01

🤔 AIの危機とバイアスの問題

AIがもたらす危機とそれに関連するバイアスの問題が議論される。アルマンは「P Doom」という概念について批判的な立場を示し、AIの安全性と将来性について考えていく。また、AIシステムにおけるバイアスの排除と倫理的な問題についても触れている。

10:02

🔍 プライバシーとAIのバランス

AIの進歩とそれに伴うプライバシー問題が議論された。個人情報の取扱いとAIの学習プロセスにおけるデータの使用について、社会がどう対応すべきかが問われた。アルマンは、プライバシーとユーティリティのトレードオフについて考えている。

15:02

🌐 AIのエネルギー消費と環境への影響

AIのエネルギー消費が環境に与える影響が話題に挙がる。AIのトレーニングに必要なエネルギーの大きさと、それが環境に及ぼす影響について話し、AIが環境問題にも貢献できる可能性についても言及されている。

20:03

🚀 AIの科学への貢献と期待

AIが科学分野に与える影響と期待が語られる。科学の進歩を加速させるAIの可能性と、それが持つ経済成長への貢献について期待を寄せている。科学者や研究者の支援ツールとしてAIが役立つとされている。

25:04

💡 AIと人間の創造性

AIが創造性と人間のアイデアをどのように支援するかが議論される。AIは人間の創造性と協力し、より良いアイデアの創出を促進する可能性があるとされている。

30:04

🛠️ AIのビジネスとスタートアップへの影響

AIがビジネスとスタートアップに与える影響が話し合う。AI技術の進歩がビジネスモデルにどのような影響を与えるか、また新しいスタートアップがどのようにAIを活用して成功を収めるかが議論された。

35:05

🏛️ AIと民主主義の未来

AIが民主主義と選挙プロセスに与える影響が懸念される。AI技術が選挙プロセスをどのように変え、それに伴うリスクと対策について考えている。

40:08

🤖 AIと教育の変革

AIが教育分野に与える影響が議論される。AIを活用した教育のカスタマイズと、学習プロセスへの貢献について期待が寄せられている。

45:09

🔮 AIの将来性と人間の役割

AIの将来性とそれに伴う人間の役割が語られる。AIが進化する中で、人間の創造性と関わり方、そして人間のユニークな能力がどう変わっていくかが議論された。

50:11

👟 創業者の旅とAIへの情熱

OpenAIの創設者であるサンジャイ・アルマンの経歴とAIへの情熱が紹介される。彼の過去の経験と、AI技術に対する情熱と期待について話されている。

🤝 人材の採用とAI企業への道

OpenAIが人材を採用する際の基準と、AI企業への道について話される。創造的思考、情熱、以及其他の重要なスキルが求められるとされている。

🚧 AI開発の進歩と人間の協力

AIの開発が進む中で、人間の協力が不可欠であることが強調される。AIと人間の相補的な関係がどのように未来を形塑くかが議論されている。

🛑 AIの限界と人間の独自性

AIの限界と人間の独自性について考えられる。AIが再現できない人間の能力と、それらが未来にどのように価値を持ちうるかが語られている。

💡 解決策の創造と新しいアイデア

問題解決のためのアイデア創造と新しいアプローチが議論される。コンテキストを変えることの重要性と、新しい環境での思考の変革について話されている。

Keywords

💡人工知能(AI)

人工知能とは、人間のように思考や学びを可能にする技術を指します。このビデオでは、AIが社会に与える影響や、MITの取り組み方、OpenAIの使命などが議論されています。例えば、AIが広範に有益であるためにMITは努力していると述べられています。

💡MITの総長

MITの総長とは、マサチューセッツ工科大学の最高責任者のことです。ビデオでは、Sally Kornbluth氏が2023年に18代目のMIT総長に就任し、多くのイニシアチブに加えて、AIの社会的貢献に注力していると紹介されています。

💡OpenAI

OpenAIは、人工知能研究と展開を行う企業であり、人類全般に利益をもたらすことがミッションです。ビデオでは、その共同創設者でCEOであるSam Altmanが、AIの将来性とその社会的影響について語っています。

💡技術革新

技術革新とは、新しい技術の開発とその普及によって社会が変貌するプロセスを指します。ビデオでは、AIが持つ潜在的な力と、それを通じて未来をより良いものにすることが議論されています。

💡バイアス

バイアスとは、特定の方向性や見方に傾斜がある状態を指します。ビデオでは、AIシステムからバイアスを除去するために必要なことや、現在のAIシステムにおけるバイアスの問題が議論されています。

💡個人情報のプライバシー

個人情報のプライバシーとは、個人が持つ情報に関する秘密保持と適切な取り扱いに関する概念です。ビデオでは、AIモデルのトレーニングに必要なデータと個人のプライバシーのバランスが探求されています。

💡環境への影響

環境への影響とは、人間の活動が自然環境に及ぼす影響を指します。ビデオでは、AIとデータセンターが能源消費と環境に与える影響が話し合いされており、AIが持つ環境問題への解決策の可能性も触れられています。

💡科学の進歩

科学の進歩とは、人類が科学的な知識を深めることによって得る利益や発展を指します。ビデオでは、AIが科学発見の速度を高めることで、持続可能な経済成長を促進する可能性が語られています。

💡教育

教育とは、知識や技能、態度を学ぶプロセスを指します。ビデオでは、AIが教育に与える影響や、個々の学習者への適応性の高い教育方法の可能性が議論されています。

💡経済的価値

経済的価値とは、商品やサービスが持つ経済的な重要性や価値を指します。ビデオでは、AIが経済システムに与える影響や、経済的価値の創出に関する議論がされています。

Highlights

Sally Kornbluth became MIT's 18th president in 2023, known for her administrative skills and advocacy for faculty and student welfare.

Sam Altman, CEO of OpenAI, discusses the company's mission to ensure AI benefits humanity.

Altman believes the 'probability of Doom' question is poorly formed and emphasizes the importance of creating a positive future with AI.

AI has evolved from a naive conception to a transformative technological revolution, according to Altman.

OpenAI has made progress in aligning AI systems with values and reducing bias.

The challenge of deciding what values and biases AI systems should have and how society defines these.

AI systems like GPT can potentially be less biased than humans and trained to behave better.

Privacy concerns are raised with personalized AI that has access to extensive personal data.

Altman discusses the balance between personal privacy and the utility of AI.

OpenAI aims to keep its AI tool widely available as a public good without ads.

AI's environmental impact is a concern, but it also has potential to assist in decarbonization efforts.

Altman emphasizes the importance of maintaining a positive outlook and striving for progress despite challenges.

Advice for young researchers to take risks, work hard, and trust their ability to figure things out.

Importance of having a guiding principle or mission statement for career decisions.

Altman's personal interest in using AI to increase the rate of scientific discovery.

Startups can leverage the current platform shift in AI to their advantage.

Warning against creating startups that rely on short-term growth without building a lasting business.

AI's impact on jobs will be significant, with some being eliminated, others transformed, and new ones created.

The need for a balanced approach to AI regulation that supports innovation while ensuring safety.

Concerns about AI's potential to influence democratic processes and the importance of preparing for new challenges.

MIT's focus on training leaders who are fluent in computer science and AI to advance their fields.

AI's potential to transform the financial sector, though specifics are not detailed.

The importance of human interaction and empathy in a world increasingly influenced by AI.

OpenAI's focus on improving cognitive capabilities and delivering them broadly and inexpensively.

Altman's personal interest in fusion energy as an alternative field of study.

Strategies for dealing with being stuck on a problem, such as changing context or seeking new perspectives.

Transcripts

thank you both for being here um they

probably neither of them probably need

an introduction but of course uh for the

record uh Sally corn became mit's 18th

president in January 1st 2023 a cell

biologist who 8year tenure at duk's

University as Provost earned her

reputation as a brilliant administrator

a creative Problem Solver a leading

advocate for faculty excellence and

student well being and in her first year

of MIT amongst many initiatives she has

focused as well on AI and has testified

mit's efforts to make sure that AI is

broadly beneficial for society uh Sam

Alman is an entrepreneur an investor a

programmer he co-founded open AI in 2015

he is their CEO and open AI is an AI

research and deployment company whose

mission is to ensure that artificial

general intelligence benefits all of

humanity Sam and Sally thank you for

being here I'm going to turn it over to

both of you great thanks so much

[Applause]

Mark so uh welcome I will say I think

this is the most uh popular event I've

seen since I arrived at MIT um and I

should say that the questions I'm asking

were uh submitted by students and sort

of gone through and curated but this

really reflects the Curiosity of our

community here and enthusiasm for you

being here so let me start with oh sorry

thank you for having me absolutely um so

let me let me just dive in so according

to columnist Uh Kevin Rose and others in

some circles the question is what is

your P Doom it's a common Icebreaker or

so I'm told most of you probably know

that P Doom or probability of Doom is

calculated on a scale from 1 to 100 and

the higher the score you give the more

strongly you believe we end up in a

doomsday scenario where AI eliminates

all human life so Sam just to break the

ice what is your

P um I I I think it's sort

of the badly formed question um hey I'm

glad I didn't take any responsibility I

I think it's like a great question for

people that uh it's a great way to like

sound smart and important and it's like

I I have as much fun pontificating on

numbers as anybody else but you know

whether you say it's 2 or 10 or 20 or 90

um the point is it's not zero um and I

think another reason I think it's a

badly formed question is that it sort of

assumes that it's a static system um I

think what we need to do is find a way

to make the future great make the future

exist like not tolerate any uh you know

branches of it we just sort of have the

the the Doom come into play but um I

think Society always

holds space for doomsayers there's value

to that um I'm happy that they exist I

think it makes us think harder about

what we're doing but I think the better

question is what needs to happen to

navigate safety sufficiently well right

so be aware as you're developing things

of that possibility but not indulge in

them too much more than aware like

really take confront it and take it

extremely seriously yes fair enough so

you know this is you've been uh in this

business for a little while now how have

your views of AI changed over the past

decade five or 10 years ago did you

expect AI GPT would become as powerful

as it

has well I honestly still think it's not

very good uh I I think we will make it

very good but we have a ton of work in

front of us

uh I think 10 years ago I probably had a

more naive conception of AI as this like

creature that was going to be off doing

stuff or this like you know magic super

intelligence in the sky that would like

figure things out and Rain money on us

and we were going to try to like figure

out how to live our lives but it was all

sort of confusing and now I think of it

much more like any other technological

Revolution hopefully the biggest and the

best and the most important and the

greatest benefits but you know we have

like a new tool in the tech tree of

humanity and people are using it to

create amazing things I think it will

continue to get way more capable and way

more autonomous over time

um but even then like I think it's just

going to integrate into society in

a in an important and transformative way

but something that is somehow going to

be you know I think like if AGI got

built

tomorrow and you asked me what would

happen the next day 10 years ago I would

have said can't really imagine it it

should be just this like absolute

transformation Singularity everything is

different all at once and I now think it

won't really be like that at all so you

think in some in some extent you were

optimistic but you couldn't have placed

yourself in the current moment in either

way and it's the same for I I don't know

like optimistic or pessimistic I think I

was just like somewhat wrong I mean

there were ways in which it was too

optimistic and too pessimistic at the

same time um you know if we make like if

we make something that is like you know

as smart as all of the super smart

students here uh that's a great

accomplishment in some sense there's

already a lot of like smart people in

the world so maybe things go faster um

maybe quality of life goes up maybe the

economy Cycles a little bit faster but

you know like if the rate of scientific

discovery becomes 10 times faster than

it is

today I I don't know how different

that'll feel to us living through it at

that time oh that's interesting that's

interesting so you know just moving

stepping back a little bit and looking

at the current models thinking about AI

systems like je chat gbt um what do you

think is NE necessary to remove bias

from the systems and can you offer an

example of thinking about bias that are

in today's AI systems and how we might

think about that going

forward I I think we've made

surprisingly good progress about how we

can align the system to behave according

to a certain set of values um I think

you know for as much as people love to

talk about this and say that oh you

can't use these things because they're

just like spewing toxic waste all the

time like if fuse GPT for behaves kind

of the way you want it to and reasonably

well and you know we're able to get it

to follow not perfectly well but better

than I at least thought was going to be

possible by this point um a given set of

values but that that gets to a now

harder question which is who decides

what what what bias means and what

values mean how do we how do we

decide what the system is supposed to do

uh how much you know how much does

society Define broad bounds around the

edges versus how much do we say

um you as a user like we trust you to

use the Tool uh you know not everybody

will use it in a way we like but that's

kind of the ca the case of tools um I

think it's important to give people a

lot of control over how they use these

tools um and even if that means that

they may use them in ways that you or I

don't always like um but there are some

things that a system just shouldn't do

uh and will have to kind of collectively

negotiate what those are I mean you know

it's interesting thinking about whether

you can make the model sort of less

biased to than we are as human beings in

a sense because you know you talk about

things like in you know in medicine uh

you know bias against certain

demographic groups for instance they're

actually trained on the way our human

doctors are behaving

right they are but then we do this rhf

step where we can exert quite a lot of

influence uh

humans are clearly very biased creatures

and often unaware of it and I don't

think that GPT 4 or five shares our same

psychological exactly flaws probably it

has its own different ones um but yeah I

think these systems can be way less bias

than

humans something to strive for you know

aside from bias other things that you

know have sort of been in part of the

public Consciousness in terms of

concerns Etc are privacy issues which

Loom large for a lot of people when

considering uh the future of llms so you

know how do we navigate the balance

between personal privacy and the need

for shared data to train AI

models I can imagine this future in

which if you want you have a

personalized AI that knows that has read

every email every text every message

you've ever sent or received has an

access to a full recording of your life

um knows every document you've ever

looked at every TV show you've ever read

every everything you've ever said or

heard or seen like all of your bits of

input in and out

and you can imagine that that would be a

super helpful thing to have you can also

Imagine the privacy concerns that that

would present and I think if we stick on

that frame and not say well should you

know AI be able to train on this data or

to that data but how are we going to

navigate the Privacy versus utility

versus safety tradeoffs or security

tradeoffs that come with that um and

like what does it even mean like do we

need a new definition of like privileged

information so that your AI companion

never has to like testify against you or

can't be subpoena by a court I don't

even know what right the problems are

going to be um but this question of

where we all will individually set the

Privacy versus utility tradeoffs and the

advantages that will be possible for

someone to have if you say I am going to

let this thing train on my entire life

um that's like a new thing for society

to navigate I don't know what the

answers will be I don't know where most

people will make the tradeoffs I don't

know what we'll say or like is even

permissible in the bounds um but we

faced a little bit of that before about

where we trade off some privacy for

utility with the services we all

use but that can go so incredibly far

with AI there are all these things that

we've had to negotiate with the internet

things about how we think about privacy

how we think about online ads um that

when you intersect them with AI become

much higher stakes and much bigger

trade-offs uh that I think we're going

to start really facing yeah I know

that's interesting in terms of um also

how much individual control there's

exerted and other words when you're

talking about aggregated data you know

good example when you're talk about

higher Stakes again is you know sort of

health record health record data Etc and

you know how much we can build into it

some sort of personal ability to sort of

set that sliding scale uh with how much

information you're willing to have as

part of that training I think in that

case we are a little

bit you know a little bit off in terms

of how we have the conversation um what

what you want out of GPT 5 or six or

whatever is for it to be the best

reasoning engine possible um it is true

that right now the way the only way we

currently know how to do that is by

training on tons and tons of data and in

the process of that it is learning

something about how to do very very

limited reasoning or cognition or

whatever you want to call it but the

fact that it can memorize data or the

fact that it's storing data at all in

its parameter space I think we'll look

back and say that was kind of like a

weird waste of resources like it is true

that gp4 can kind of act like a database

but barely it's slow it's expensive it

doesn't work very well it's not it's not

really what you want um it's just kind

of as a side effect of the only way we

know how to make a model that a

reasoning engine right now um it has all

these other properties but I assume at

some point we'll figure out how to sort

of separate the reasoning engine n from

all of the uh you know needing tons of

data or storing the data in there and

we'll be able to treat them as sort of

separate things and I think that'll make

some of these privacy issues easier no

that makes a lot of sense um speaking

about openness there's been a

considerable discussion of whether open

AI is actually open you said that while

it may not be completely open source

it's open in other ways can you say a

little bit more about this and how you

think about it um we make a great free

AI tool available hundreds of millions

of people hopefully billions of people

will use it in the future uh we don't

run ads we just do this as like a public

good because we think it's important to

put the tool in people's hands um and we

want it to be very widely available very

easy to use very helpful

um I think that's just a cool thing it

is a cool

thing you know it is funny how um how it

starts as you're talking about how AI

gets built into a or our normal lives

and that'll evolve I mean I think

probably most of us remember the first

time we saw chat GPT and we're like oh

my God that is so cool now we're trying

to think about what the next Generations

you know what are the next generation of

coolness going to be by the way I think

that's great I think it's awesome that

for um you know 2 weeks everybody was

freaking out about gp4 and thought it

was super cool and then by the third

week everyone was like exactly come on

where's GPT 5 I'm tired of waiting EXA

exactly I I actually think that says

something like legitimately great about

human expectation and striving and why

we all have to like continue to make

things better um and I think it's like

great that a baby born today uh will

never know a world in which the products

and services they use are not

intelligent we'll never know a world in

which cognition is not like abundant and

part of everything that you use so I

think this like this human discontent

with the state of things and the

expectation that the world should get

better every year uh I think that's

awesome yeah no great I agree um so in

the past year uh new electricity

speaking of less exciting edge of things

the new electricity demand from Ai and

data centers has been cited as an

environmental concern at the same time

many you know talk about AI assisting

and decarbonization what are your

thoughts about this the tension between

its effect on climates climate its

ability to potentially help us uh fight

the impact of climate change uh as it

moves

forward I'll answer that specifically

and a more General observation um it is

true that AI needs a huge amount of

energy but not huge relative to what the

rest of the world needs if we have to

spend and I I don't even think you know

if we spent 1% of the world's

electricity training powerful Ai and

that AI helped us figure out how to get

to uh non-carbon based energy or do

carbon capture better that would be a

massive win um and even if we didn't do

that uh if that 1% of compute that we

spent on AI let people live their lives

better and have to

like yeah I read this thing about the

compute Google used once compared to the

amount of carbon that people used to

spend driving in their cars places to

get information and you know you have

people saying Google's so horrible we

should shut it down it's like spending

El energy and it's it was intellectually

a very dishonest thing to say because it

was a net Savings in energy uh I think

pretty clearly the internet in general

and the you know what it lets us do for

telecommuting probably also a savings um

so you know for like AI yeah it's going

to need a lot of energy we're going to

keep figuring out way more efficient

algorithms way more efficient chips

we're going to get Fusion we're going to

power the stuff this way um so I think

it is like important to address this

issue um but we will in all of these

fantastic

ways but I think this points to

something else which is the you know you

open asking about P doom and the level

of doomeris in society right now I think

the way we are teaching our young people

that the world is totally screwed that

it's hopeless to try to solve problems

that all we can do is like sit in our

bedrooms in the dark and think about how

awful we are is a really deeply

unproductive streak and I hope MIT is

different than a lot of other college

campuses I assume it is but you all need

to like make it part of your life

mission to fight against this Prosperity

abundance um you know a better life next

year a better life for our children that

is the only path forward that is the

only way to have a functioning society

and there will always be people who want

to sit around and say we shouldn't do AI

because we may burn a little more carbon

or we shouldn't do AI because you know

we haven't fully addressed bias and it

turns out a couple years later we made a

lot of progress on both of those things

and the anti-progress streak the anti

likee people deserve a great life streak

who are usually the people that have

quite a lot of privilege in the first

place um is something I hope you all

fight against God yeah

[Applause]

you know I I uh you know coming to MIT

fairly fresh from the outside I think

this is sort of core to the MIT ethos

which is you know naming the problems

and figuring out a way to solve them I'm

very happy to be here yeah it's

fantastic we're very happy you are here

the other thing you said that really

struck me as we think about the costs of

AI I don't just mean monetary costs

whether it's climate or anything else it

is this notion of

deducting what the uh what AI can

contribute to the problem as not a c you

know what I mean as sort of a long-term

term uh balancing of The Ledger that I

think is important yeah so that's

interesting um does open AI intend to

build tools that will specifically

impact science and engineering or will

you be more focused on sort of business

and consumer

applications um for sure we intend to do

that I think the most like personally

the thing I am most most interested in

is how we use AI to increase the rate of

scientific discovery I believe that is

the core engine of human progress and

that it is the only way we drive the

sustainable economic growth that we were

talking about earlier people aren't

content with gp4 they want gp5 they want

things to get better um everyone wants

like more and better and faster uh and

science is how we get there so of all of

the things of all the great things that

AI will do um I am personally most

passionate about the impact that I hope

it will have expect it will have on

science that said um it may this may all

be more one-dimensional than we think um

if we make a great AI tool that can help

people solve any kind of problem in

front of them that can help people

reason in new ways uh that's great for

consumers that's great for scientists

that's great for businesses it's great

for Education it it may the the G of AGI

the general is sort of the surprising

piece yeah I know that's really

interesting and now I come back to your

comment about you know sort of getting

in your old car and driving to the

library you know you know a lot of that

the the creativity part is still human

but a lot of the um aggregating all of

the knowledge that you that you can use

as a launching Point can really be

expedited you know by asking a few key

questions uh to AI up front totally I

mean again the the what any one

individual and certainly what any group

of us will be capable of um I think it's

going to if we could go see what each of

us can do 10 or 20 years in the future I

think it would astonish us today yeah um

if you know it's like maybe in a few

years it's like each of us has like a

great Chief of Staff or uh like PhD

student or whatever analogy you want

that's off like helping us optimize

ourselves and do our best work and our

best ideas and whatever and then maybe

at some point it's like each of us has

like a full company full of like

brilliant Experts of anything um just

working super productively together cool

so you know what do you have what advice

do you have for we have a lot of young

researchers or people who are aspiring

to be young researchers in the audience

uh what's sort of your general advice uh

for making a real impact in the world

and you know you alluded to it in terms

of thinking about possibilities and not

sitting in your bedroom in the dark I

think that's a good base recommendation

um but I'm just wondering what else uh

what else you might want to say to this

audience about that um first of all I

think this is probably the most exciting

time to be launching your career um in

many decades maybe ever I don't know but

it's like whatever it is it's a really

big deal and the fact that you have this

huge Tailwind means I think you can um I

think you can take more risk than usual

I think if you do something doesn't work

out there's just going to be phenomenal

opportunities for a long time uh I think

you have you can have more impact than

normal and so there's like a premium on

you know having this be a period where

you work really hard I certainly would

be biased to do something with AI um but

like of course I'm going to say that so

maybe it's wrong

um I think in general the the the kind

of

core the the most important to lesson to

learn um early on in your career is that

you can kind of figure anything out um

and that no one has all of the answers

when they start out but you just sort of

like stumble your way through it have

like a fast iteration speed try to like

drift towards the most

interesting problems to you and be

around the most impressive people and

have this like trust that you'll

successively iterate to the right thing

and and you can kind of like you can do

more than you think faster than you

think uh and people it takes a while to

learn that lesson um but it it it it

gets you know you see it work a few

times and you really start to trust it

um and so like you can just do stuff

sounds like not real advice or like very

empty advice but I think is like it's

it's much more profound than it sounds

on the surface the other thing I would

say is figuring

out relatively early on and this takes

some practice kind of like what your own

personal I don't even know what to call

it like passion missionstatement like

the kind of way you want to spend your

time or what you really care about um

and we talked about like this concept of

like techno abundance as a way to

drive um like prosperity and better

lives for people that that's been

something for me that has always really

resonated and I've always tried to

figure out like how to work on that but

having some sort of like letting letting

yourself develop some sort of like

guiding principle of how you make

decisions about how to allocate your

time and where to try to like steer

things that that was like that's been

very helpful to me yeah I think this

follow follow your passion and also you

know from you're painting a picture of a

world where there's sort of infinite

possibilities and you know doing you

can't always be so strategic about what

you think is good for you to do you want

to do something that imagines all those

possibilities and follows those passions

it yeah for me it's like like passion is

not

quite the right word it's like something

closer to like what is the moral

obligation for me to work on and then

and I on like the really bad days is

when I'm not having fun that's somehow

like it's much more motivating than just

the thing I like doing the most that's

interesting um another element of this

and this is actually a really core part

of uh mit's culture is entrepreneurship

and so we have a lot of aspiring

entrepreneurs so there's you know

developing the sort of underlying ideas

but how do you think about building um

successful companies in today's

ecosystem and what part of the value

chain should new where where should new

startups sort of focus their effort

again I think this is like the best time

for new startups in particular in a very

long time uh startups tend to succeed uh

right around the time of big platform

shifts big companies are slower and less

Innovative than startups but they have a

lot of other advantages um when you get

the speed and iteration and cycle time

Advantage the most is when like the

ground is shaking

um and right now I think you can there

was like a moment like this right when

the internet happened there was a Moment

Like This although smaller after mobile

uh there was another moment also smaller

after AWS and this idea of like cloud

services and then for a very long time

like more than a decade we've just been

sort of waiting and I think now we

finally have a new platform and so if

history is a guide which usually it is

and I suspect it will be this time um

it's an amazing time to start a company

and the advantages you have as a company

are you can move much faster you can

like live in the future more than like

big companies that have quarterly or

annual or whatever they have planning

Cycles um and that's how you win and I

think this is a great time to do it

excellent I think there's a lot of

people here who uh are take that to

heart can I say one more thing about

that um I will issue like a that was all

the positive um here

here's a warning

um with any new tech platform you can

always drive phenomenal short-term

growth and so you have this class of AI

startups like you used to have a class

of mobile startups and before that like

you used to have a class of Internet

startups that were not building an

enduring business um but instead we're

building this sort of like novelty thing

that was and be and you kind of delude

yourself because you get amazing fast

growth

and because there's this like magic new

technology and you know the dust hasn't

settled yet just because there is a

magic new technology it does not excuse

you from like the laws of physics of a

business you still have to figure out a

way to build um some sort of switching

cost some sort of relationship with

customers some sort of compounding

advantage over time and in

the in the Gold Rush moments I think

startups at their Peril often forget

that so you still have to like do all

things a business always has to do yeah

that's really that's I think important

important advice um you know the other

thing you know it's interesting how this

question was phrased and I'll read it in

a minute but now that I've heard you

talk a little bit I might phrase it a

little bit differently the question was

phrased in what ways might technology

like chat GPT threaten versus help the

future of work but uh it sounds like you

you tilt much more towards help but also

thinking about you know what that means

in in real terms how does it how does it

help people in their future employment

one of the things that annoys me most

about people who work on AI is when they

stand up and with a straight face say oh

this will never cause any job

elimination yeah you know this is just

an additive thing this is just going to

it's all going to be great like this is

going to eliminate a lot of current jobs

and this is going to

change the way that a lot of current

jobs function and this is going to

create entirely new jobs

it that always happens with technology

um it's probably never happened this

fast although again we may be like

drinking the Kool-Aid too much and the

inertia of society may be such that it's

slower than we think

but I kind of expect we're only a

generation or two away from models that

for the first time show some degree of

real economic impact good and bad um but

something you can measure and there will

be classes of jobs that totally go away

there will be classes of jobs where you

have to change what you do a lot there

will be classes of jobs where the

productivity compensation whatever you

want to talk about whatever measure goes

up by like a giant Factor um and then

there will be things that feel like jobs

to the people of the future that to us

today look like a complete Indulgence

and waste of time as what many of us do

today would look like to people from

hundreds of years ago uh

I think as long as you believe that

humans very deeply want

to create and be useful and feel like

they're making like relative

differential progress all of which are

things I would bet on hard um we're not

going to run out of things to do

um I love reading contemporaneous

accounts from people living through

previous technological revolutions and

what they say about man we're all going

to only work four hours a week if we

work at all and we're just going to you

know it like you say it every time

it in some way it does feel to me like

this time is

different and as a matter of degree it

might be and as a matter of speed I

really think it will be um and I have

some concern about how quickly we can

adapt to this kind of

change but I have no real concern that

we can eventually adapt to this kind of

change um I'm sure the social contract

will change um I'm

sure most jobs will be different in the

future than they are today but like the

Deep human drivers don't seem to me

likely to go anywhere interesting and

and obviously different categories of

jobs are going to be really

differentially affected for sure um so

uh with President Biden's recent

executive order on AI as well as the

Congressional hearings on AI regulation

um there is a concern that regulatory

Frameworks might solidify the position

of established players might stifle

Innovation competition accessibility how

do you envision AI be regulation because

we really are at a critical moment being

designed to uphold Innovation and

competition while ensuring that the

field remains accessible for merging uh

players to Pioneer the next

transformative

Technologies um I I think we faced

versions of this with other kinds of

Regulation like you want to know that

the food you buy in a g grocery store is

unlikely to make you sick and we kind of

all agree that regulation there is good

but you also want to be able to like

grow food in your backyard without

having to like go through a bunch of

like hoops and you get to do that too

um I think for AI systems there will be

some threshold above which we say okay

uh the system presents a level of risk

that we don't want to take without

reasonable safety precautions and then I

think there will be a level of AI

systems where we say even though there's

going to be misuse um we should open

source these and let people use them and

there should be no regulatory burden on

companies developing them um because

we're willing to make the Innovation and

freedom

tradeoff for the negative safety

consequences at level X right and then

level y can be totally different

and I totally get the impulse to say

any regulatory action is unacceptable

because it's just big companies that are

going to use it to like for regulatory

capture um and you know if Society

decides we don't want to regulate AI at

all and we'll just take our chances I'll

accept the outcome of a democratic

process it seems to me good to have some

voices saying like well you open with a

p Doom question exactly um if if the

framing of that question is

correct uh then it seems to me useful to

have some voices

saying let's not act out of fear but

proceed with some reasonable caution no

that makes sense um speaking of which um

you've said you believe that the

upcoming presidential election won't be

the same as the last one I think we all

think that uh but there are lessons to

be learned from 2020 uh what are these

lessons how can we mitigate the risk AI

Pro opposes to the Democratic process

and to uh the future of Democracy in

America you

know maybe maybe the use of advanced AI

will be the least interesting thing

about this election the way it's shaping

up uh

I I do

think yeah I think there will be like

better deep fakes of course and there

will be better like troll Farms of

course um

what I think is more interesting is

trying to get ahead of to the degree

that we can this is easier said than

done but trying to get ahead of the new

things that just weren't possible before

um so like customize one-on-one

persuasion where uh an AI system reads

all of your social media posts and

targets something just at you that

wasn't really possible with all of the

sort of like online disinformation and

trolling of the last election right and

that's the kind of new thing that I wish

we were taken more seriously yeah I mean

that's the logical extension as you

mentioned of you know advertisements

coming up when you're surfing the web

people know you buy X kinds of shoes or

whatever how you think about things

absolutely um a more local question uh

one of mit's educational priorities is

to train tomorrow's leaders to be in

essence Computing bilingual meaning that

regardless of their chosen field they

will need to be fluent in computer

science and AI to advance their work

work can you comment on the impact of

our way of thinking I don't know if

people talk you about blending Blended

Computing thinking about how we CH train

bilinguals how they can start thinking

about the future

careers really learning two different

fields and using Computing as a way into

into other

areas one of the general observations I

make about the history of computing it

has it gotten

increasingly more accessible and more

natural over time M um

you know like I heard these stories of

people with like punch cards that they

would like have these crazy systems

about how they sorted them and they

would drop them and that was a big mess

and it was like that was like a not a

natural thing to use

um undergraduate education sounds tough

um low-level programming languages were

a step forward um but still like not a

thing that most of the world knew how to

use or I think a very natural tool in

some sense um and then you get to like

current programming languages and

they're way more accessible and way

easier to use also way more expressive

and more powerful

um along with that Evolution you go from

uh like the command line um to a gooey

to like a mouse and a keyboard to just

like touching your phone like you don't

that's that was like not very that was

pretty natural right um and then you go

to language which is Supernatural um you

know like people are very good at

language as a way to use the computer

you can sort of ask chat GPT something

but also way to program the computer so

you saw these things like converge to

this one interface

and you don't have to be that bilingual

anymore right like right the the you get

like the same way that you like talk to

a a friend or a colleague you can talk

to a

computer and this is I think a more

profound thing than it sounds like on

the surface um the degree to which we

can push

Ai and people to have the same kind of

interface um so I'm like more excited

about humanoid robots than I am about

other forms because I think the world is

just very designed for humans and we

should keep it that way but we want the

benefits of robots that can help us um I

think we want AI systems to do their

cognition and language to communicate

with us in language um it's a very

human Focus thing um so my hope is we

don't all have to be a bilingual that's

interesting I mean because we have I

would

say an over an overwhelming number of

our students are obviously interested in

CS and now that we've sort of rolled out

you know these Blended areas I think you

know it may be sort of the sort of first

generation of what you're talking about

in other words people who who then you

know next Generations it will be

completely intuitive totally yeah so

that's really interesting um let's see

uh there's there's some uh backup things

here that other folks subit so how do

you think AI will impact the financial

sector thinking about sort of banking

and Equity have you thought about that

at all um first of all if we're if we're

on the backup questions and we want to

like let people shout out if you want to

do that I'm totally down otherwise we

can do them either way let me see we can

we can do some uh well let me ask you

then well why to answer that I have a

couple other things and then then if

people really feel compelled to yell out

some questions related to this interview

um then then

then we can do that but go ahead

um

I I haven't thought as much as I would

like to about any specific area cuz

figuring out how to get the general

purpose intelligence and what that means

has been pretty allc consuming um you

know like education and Healthcare are

kind of maybe the two specifics that

I've thought about the most on something

like the financial system uh I expect AI

to like impact that kind of as much as

everything else but I don't think I have

like a deeply insightful specific thing

to say about here's how this really

transforms let's think a little bit more

than about the educational sector you

know we think a lot about how um you

know people worry about things about AI

in the classroom but I really think

there's just a huge potential for how we

teach and how we tailor things to

individuals Etc so if you could say a

word about that I think that would be

great

I think with what you're seeing people

do already just with like regular gbd4

in chat gbt when you say like please

pretend you're a tutor and help me learn

this thing if that can work so well then

as people start to take Next Generation

models and customize them for learning

experiences we're we're going to be in a

very good place um and it's it's like

pretty awesome to see what people are

building already um it's great to hear

from teachers about the impact it's had

on their students it's great to hear

from students about what they're

learning on their own but this one seems

like a kind of Slam Dunk good use no

absolutely so this isn't this isn't one

of the sort of pres submitted questions

but I'm always really interested when I

meet people who are doing really

interesting things how you came to this

in other words what was your sort of

path and how did you how did you think

about and make choices as you came to

what you're

doing um well I was like a very nerdy

kid uh I spent like a lot of time

reading sci-fi or watching Star Trek or

whatever and uh I

like that was you know that was like the

that was already a time where all

contemporary sci-fi was pretty uh

dystopic but the older stuff the older

Star Trek episodes those were still like

kind of optimistic and cool and you sort

of like saw how AI was going to be great

and kind of how that was obviously the

future um and I kind of like

abstractedly thought like I would love

to work on that someday that would be so

cool I never thought it would actually

happen um but I was always interested uh

and then like you know life went on I

got a little bit older uh I went to

school and I decided to study in the AI

lab

um I also got really interested in

energy um but this was kind of when I

started to believe that like if we could

use tech technology to deliver abundance

that would not solve the elb problem

people would still find all sorts of

ways to be unhappy but it was like

something I really wanted to do um

and uh worked in the AI lab it was like

clear that AI was nowhere close to

working it's like back in 2004 um got

sidetracked a bunch and went off and did

a bunch of other stuff and you know did

a startup became a startup investor and

then in 2012 uh noticed the Alex net

paper took me a couple of years to

really internalize maybe this thing

finally working and I was like should

really do something

here excellent now I think there's

probably one some question on the minds

of some of the people here uh what do

you look for when you hire somebody I

mean there'll be people sitting in this

audience who are wondering that

intensely first of all we would really

love to hear from you and I think this

is probably the most the most or second

most exciting time in open a history

right now um so uh this is like this be

a good time to give us a call

um number up on the screen what do we

look for

uh kind of all the obvious stuff like

original thinkers smart driven dedicated

people people who sort of like are

particularly driven to work on AI versus

you know there there's always like a set

of people who will go work on whatever

like the hot area is uh and then there's

like a set of people who are like no

this is like the thing and we like

people for whom this is their thing um

but that's kind of it nothing I mean all

of the obvious things and nothing that

unusual interesting if you were doing

something else what would it be I would

go work on Fusion full-time Fusion was

going to be something else yeah there's

definitely people here who are

interested in that as well um you know

we do have a few minutes um yeah so we

don't have any mics I see monus's hand

up right away uh but you got to shout

I'll repeat the question hi first of all

thanks for all all you're doing to the

world

andit um so I have some of the smartest

kids in the world even smartest myam are

coming to me and saying um how do I how

do I want to plan my time for the next 5

years while humans are still helpful and

useful and it it's it's striking because

you know I see AI as an extraordinary

tool and I am so looking forward to the

future of work where I can do amazing

things that I can't do right now where

all the mundane stuff that I'm doing now

is replaced and I can do super awesome

new things but there will come a time

given the two curves of how fast human

intelligence has progressed through the

old school Evolution and nurture of

course and the curve with which AI is

progressing which is just ridiculous so

these curves are going to cross very

soon and I'm just curious what are some

unique capabilities that humans have

that you think we will not be able to

replicate with AI what are some of the

architectures that will help us maybe

replicate those things and push beyond

that and how do you see the future of

human versus machine intelligence more

broadly two conflicting answers to that

um number one I kind of suspect

that forever onwards from now it's

always going to feel like man this next

5year period is so critical this is when

I can really contribute and after that

who knows what happens but in practice

there's always going to be um it's

always going to feel like man this curve

of intelligence is rising so fast right

now I can use these tools to like out

aieve but eventually they outrun it but

at any point on that exponential um

you'll always be able to use the tools

to do amazing things and it'll always

feel like then you get totally outrun

but you never quite do because like we

just become more capable um and I didn't

used to really think this way I'm still

not sure I'm right uh but it does seem

to me

like we will just be able to like

accomplish more do more things we'll

have ridiculous the expectations will go

up too to like participate in the

economy in some sense but what we can

get done

uh before like we feel like this wave is

going to crash over us we'll always feel

like man in this 5 years this is my WI

but it's going to be a rolling 5 years

forever um and I think we'll find that

humans are really good at like humans

are we're just we're so wired to care

about other humans we're so wired to

like focus our energy know what other

people want be focused on like

delivering value for others uh that you

know like I can I can see very extreme

worlds where like human money and

machine money are just different things

um but but there's like an increasing

premium on human the human money part of

that of That World um I don't know

exactly what it's going to look like but

I do kind of believe in the biological

drivers of humanity not changing that

much uh and then the answer on the other

side

is and this is just speaking personally

this is not a well-reasoned or defens

like logically defended thing at all

this is just what I feel I do feel like

like I know I'm going to be nostalgic

for this

time and it's sort of a strange thing to

like feel that while you're living

through

it that's really

interesting I see another hand back

there well I'm going to have to ask you

to just hand the microphone to whoever

you see because I

uh can you hear me here sure

more bullish on startups building and

user applications or infrastructure

uh probably at this current phase of

where we are on end user applications

you got to like pick the right ones you

got to pick the ones that that benefit

from the models getting better not the

ones that are betting the model actually

doesn't get better and kind of are

fixing the current generation proc s

but I think it's

like I think there's like a lot of value

to unlock there um building

infrastructure can be great too I think

you can really succeed both ways but

like the number of what feel like 100

billion dollar application layer

companies right now where you can really

like do something incredibly useful to

people quickly seems pretty exciting to

me yeah

uh

yeah thank you so much for giving this

talk I just had a question in terms of

the future that you see specifically for

open AI I was just wondering you can

come out with GPT like 6 7 8 9 10 20 30

40 50 60 um you have like your creation

of Sora um a lot of companies are

creating like life-size robots that are

utilizing AI as the backend or a lot of

technological companies also using chat

GPT as like a foundation for their own

llms so I was just wondering for you

guys specifically at open AI what do you

guys foresee as your niche in the future

once you start I guess not perfecting

but getting to near Perfection for like

your llm

models I'm going to go over there um I

think it's so far

we're so far away from when we start to

level off

um that that's not currently on our mind

um I think you know at least for the

next three four model

Generations I believe we can make it so

incredibly much better every time um we

should focus on that and if we can do

that everything else will kind of work

out uh like I think our our Niche or

whatever is we want to deliver

great useful impressive cognitive

capability for as uh abundantly and

inexpensively as we can and there's

other good things we can do but if we

can just focus on that I think it'll be

like a great service to the world and we

can go on that for a

while uh I have two questions one where

did you get your

shoes uh Adidas did a Lego collaboration

I love

Legos amazing cool um my my serious

question is how soon do you predict that

we'll get to artificial general

intelligence and what is open ai's role

uh in getting us there you know I don't

I don't I I I no longer think there will

be a time where the world agrees okay

this was the year we crossed the AGI

threshold I think the the phrase has

become

so uh so overloaded as how as a

definition I think there are people who

would say we'll get there soon I think

there's like a lot of people who will

say by 2040 like when we have these

unbelievably capable systems ah it's not

quite AGI yet I can't do this one thing

um so I think I think the question is

like the only way I know how to like

form the question well at this point is

what is the range of time that we get to

like capability x y and z um but I think

that the AGI like I can't make myself do

this cuz it's like I it's too much of

like a it's too like deep in the OS at

this point but I like I try to not use

the word AGI anymore it's a total like

I'm never going to succeed at that but

it's like you know when can we when can

we do that new scientific discovery in

some areas when can we uh when can we

like add a lot of economic value um I

don't know I expect that like by the end

of this decade we have systems that

create really significant economic

value excellent so maybe last question

uh from the audience where whoever's

last one here maybe and then uh I'll

close this

out

yeah thank you so much Sam for the talk

um I guess I have more a general

question I guess what do you do when

you're have a problem to solve and

you're stuck and you don't know the way

forward um guess what's your thought

process on trying to get unstuck I

somehow try to like change context I try

to talk to different people about it um

I try to like in an extreme case I'll

like go travel somewhere to like really

like kind of change things like the one

time I like jet lag is if I'm like

really stuck on a problem and I wake up

in the middle of the night in some new

context and it seems to be helpful um

but like some way or other I try to

change

context well with that um I really want

to thank you that was fascinating I

think everybody greatly enjoyed the

conversation thank you all very much and

[Applause]

uh thank you thank you

awesome sure

can I sign this real

[Applause]

quick he went that