【スターゲートがAIの利用爆増を支える】オープンAI・アルトマンCEO/トランプ氏は直感で良い判断を下す/1時間で100万ユーザー増「ジブリ風」画像生成で浮上した課題/発端はサプライチェーンの限界
Summary
TLDRこのインタビューでは、AIの急速な進化とそれに伴う計算能力の必要性について、OpenAIのCEOが語っています。特に、Stargateプロジェクトの立ち上げや、AIモデルの需要の急増が強調され、5000億ドルの資金が必要な理由が説明されています。また、AIが雇用に与える影響や、科学的発見を加速させる可能性についても言及されています。技術の進化が新たな雇用を生む一方で、既存の職業が消失するリスクについても触れられています。
Takeaways
- 🚀 AIの需要が急増している
- 💰 5000億ドルの資金が必要
- 🔍 AIは科学的発見を加速する
- 🤖 AIは新たな雇用を生む可能性がある
- ⚙️ Stargateプロジェクトは大規模な計算インフラを目指す
- 📈 技術の進化により効率的なチップが開発される
- 🌍 AIの競争は激化している
- 📊 計算能力の不足が課題
- 💡 AIは人類の理解を深める
- 🔗 Deepseekの効率的な電力供給方法に注目
Timeline
- 00:00:00 - 00:05:00
このセクションでは、AIモデルの需要が予想以上に高まっていることが強調され、Stargateプロジェクトの必要性が説明されています。特に、GPT-4のリリース後、計算能力の限界に直面し、より多くのコンピューティングリソースと接続性が求められるようになったことが述べられています。
- 00:05:00 - 00:10:00
次に、Stargateプロジェクトの資金調達とその背後にあるパートナーシップについての詳細が語られています。特に、MicrosoftやSoftBankとの協力が重要であり、5000億ドルの資金が必要な理由が説明されています。AIの需要が急増している中で、OpenAIが持続可能なビジネスモデルを確立できるかどうかについての自信も示されています。
- 00:10:00 - 00:17:42
最後に、AIの進化がもたらす雇用の変化についての懸念が取り上げられています。AIは多くの仕事を奪う一方で、新しい仕事を生み出す可能性があることが強調され、技術の進化がもたらす影響についての認識が求められています。特に、AIが科学的発見を加速させる可能性についての期待が語られています。
Mind Map
Video Q&A
Stargateプロジェクトとは何ですか?
Stargateプロジェクトは、AIモデルの需要に応えるための大規模な計算インフラストラクチャーの構築を目指しています。
なぜ5000億ドルの資金が必要なのですか?
これは、今後数年間の成長予測に基づいて必要な計算能力をカバーするための資金です。
AIはどのように雇用に影響を与えるのですか?
AIは既存の職業を消失させる一方で、新たな職業を生む可能性があります。
OpenAIはどのようにして持続可能な企業になるのですか?
現在の成長と収益性の見込みに基づいて、OpenAIは持続可能な企業になると考えています。
Stargateはどのように設計されていますか?
データセンターの設計は、計算需要に基づいて最適化されています。
AIの未来についてどう考えていますか?
AIは科学的発見を加速させ、人類の理解を深めると信じています。
Deepseekの効率的な電力供給方法についてどう思いますか?
Deepseekのチームは優れた成果を上げていますが、私たちの方法も効率的です。
AIの需要はどのように変化していますか?
AIの需要は急増しており、計算能力が不足しています。
AIの進化はどのように進むと考えていますか?
技術の進化により、より効率的なチップやアルゴリズムが開発されるでしょう。
AIの競争についてどう考えていますか?
私たちは日々改善を目指しており、競争に対して最善を尽くしています。
View more video summaries
【仕事は食われるがコンサルは足りない】アクセンチュアAIセンター長・保科学世/「リスキルが必要な社員39%」もAIが算出/プレゼン資料を“秒”で作る極秘AIをデモ公開【CROSS DIG 1on1】
2025デザイン情報学I ブルーの歴史
ウクライナから日本に来て3年 「架け橋になりたい」祖国と家族を思いながら夢を追う女性の奮闘記【テレメンタリー】
上達が遅くなる考え方 (ステップをちゃんと踏もう編)#ドラム練習の仕方
【スト6】モダンブランカでマスターまでにやった事とやらなかった事【ストリートファイター6】【モダンブランカ】
【徹底解説】NTTの未来戦略で株価大変動! IOWN構想と住信SBIネット銀行の買収で何が起こるのか?
- 00:00:00[Music]
- 00:00:03Hello. Good to see you. Thank you.
- 00:00:07So, I saw what's happening in Abalene
- 00:00:09with my own eyes. Isn't it cool? It's
- 00:00:11happening. Take me back to the beginning
- 00:00:12before Stargate. What did you start
- 00:00:14seeing that made you realize we need
- 00:00:16more compute, more power, more
- 00:00:18connectivity that you were hitting a
- 00:00:19limit and needed to scale?
- 00:00:22We used to think a lot about the compute
- 00:00:23we would need to train the models. And
- 00:00:27what we didn't used to think about as
- 00:00:29much is how much people were going to
- 00:00:30use these models. Um I mean we did think
- 00:00:32about it. It just turned out people want
- 00:00:33to use the models much more than we
- 00:00:36imagined. And a couple of years ago
- 00:00:39maybe after the launch of GPT4 in Chhat
- 00:00:42GBT it really started to hit like oh man
- 00:00:45this is this is like gone from a lot of
- 00:00:47compute to like the biggest
- 00:00:48infrastructure project in history. Yeah.
- 00:00:50And we started thinking about ways to do
- 00:00:51it and we started trying to understand
- 00:00:52where the limits in the supply chain
- 00:00:54were and
- 00:00:57and out of that emerged Stargate. Um
- 00:00:59there were other smaller things that we
- 00:01:01did first. We have you know worked with
- 00:01:04uh other partners. We've worked with
- 00:01:05Microsoft to build out like very
- 00:01:07gigantic amounts of compute. But um you
- 00:01:10know this is like the next step in that
- 00:01:11evolution, right? I mean you traveled
- 00:01:13around the world talking to people about
- 00:01:15this. How did you get to talking with
- 00:01:17Masayoshi's son and Larry Ellison and
- 00:01:19how did you all come together? Um I'm
- 00:01:23well I've known Masa for a long time. Um
- 00:01:25but I I was on a in 203
- 00:01:31I did sort of two long trips. I think
- 00:01:32that was the year that's all it's all
- 00:01:34about. Um I I did two kind of long trips
- 00:01:37around the world and a lot of it was to
- 00:01:39talk to developers and governments. Uh,
- 00:01:42but a lot of it was also to just really
- 00:01:44try to like get my head around the
- 00:01:46supply chain. This is not something I've
- 00:01:48done before. Like I had not before
- 00:01:49thought about like what it was going to
- 00:01:50take to get compute up and running at
- 00:01:52this scale. And there are a lot of like
- 00:01:54hard pieces. Um, and on one of those
- 00:01:58trips, uh, I met with Masa. Um, Masa has
- 00:02:02like thought for a long time about, uh,
- 00:02:05chip fabs in particular, but really kind
- 00:02:07of the whole thing that it takes. and we
- 00:02:09got to speaking about what it would um
- 00:02:11you know what it would take to do
- 00:02:13compute at this kind of scale and then
- 00:02:15it took us a while to to figure out I
- 00:02:16mean it is a is a complex supply chain
- 00:02:19with a lot of partners and obviously a
- 00:02:20lot of capital and you know Soft Bank's
- 00:02:22obviously your financial partner Oracle
- 00:02:23is the technical partner why couldn't
- 00:02:25you get what you needed from Microsoft I
- 00:02:28mean we do get a lot of great stuff from
- 00:02:29Microsoft but I think this is more than
- 00:02:31any one company can deliver mic we have
- 00:02:33Microsoft will do a lot of compute for
- 00:02:35us a lot we're very happy about that why
- 00:02:38call Stargate. Um, it began as a code
- 00:02:42name and sometimes code names stick,
- 00:02:45but it means something, right? Oh. Um,
- 00:02:48the design of one of the very early
- 00:02:52layouts of a data center looked a little
- 00:02:53bit like Stargate from the from the
- 00:02:55show. Mhm. A futuristic wormhole. Not
- 00:02:58like that at all. Just Okay. You
- 00:03:00announced this at the White House the
- 00:03:02day after the inauguration.
- 00:03:04How did this get to President Trump's
- 00:03:05desk and and what was that moment like?
- 00:03:09Well, the president is super interested
- 00:03:10in infrastructure and he's made uh you
- 00:03:13know a big priority which I think is
- 00:03:15wonderful about uh permitting the energy
- 00:03:20production and the facilities for the
- 00:03:22data centers. Uh so that was kind of how
- 00:03:23the conversation started. Okay. We're in
- 00:03:26the middle of a step change for AI and
- 00:03:28clearly what we saw just in the last
- 00:03:29week from you alone. um you're seeing
- 00:03:32what's next on the road map. How does
- 00:03:34that inform what the design of the data
- 00:03:38centers of the future need to be?
- 00:03:43I
- 00:03:44I mean I think there's like a lot of
- 00:03:46details we're learning about how to
- 00:03:47design these data centers, what you want
- 00:03:50um really at all levels from like the
- 00:03:52chip to the sort of architecture of the
- 00:03:54whole data center.
- 00:03:56But but the main the main thing that's
- 00:03:58been on my mind and I think on many
- 00:04:00people's mind is just how much inference
- 00:04:02demand there is. Um we are crazily
- 00:04:06constrained. We have a gigantic comput
- 00:04:09fleet like gigantic gigantic and yet
- 00:04:12still um if we had twice as much we
- 00:04:14would be able to offer much better
- 00:04:15products and services and there's just
- 00:04:16no shortage of demand. So
- 00:04:20for me, there are all the technical
- 00:04:22lessons about what we've learned and how
- 00:04:25we want to build this, but mostly we
- 00:04:26just want a lot. Explain the math to us.
- 00:04:29How does it all add up to needing $500
- 00:04:31billion? Because folks out there don't
- 00:04:33get the accounting, who's putting in
- 00:04:36what, how much funding is really
- 00:04:38secured. Um, how does it add up in terms
- 00:04:40of like what do we need that for? How
- 00:04:42how does the funding come together? Like
- 00:04:43explain that number 500 billion. Oh. Um,
- 00:04:47well, that covers the capacity we think
- 00:04:49we need for the next few years given our
- 00:04:50growth projections. Uh, you know, we we
- 00:04:54the the interesting question is if we
- 00:04:56had if we knew how to get a trillion
- 00:04:57dollars right now, which we don't. Um,
- 00:05:00would we be able to deploy that
- 00:05:02profitably in the next few years?
- 00:05:05And I'm not sure about that, but I feel
- 00:05:07confident we can like make 500 billion
- 00:05:08of value back. Let's talk about that
- 00:05:10because you just raised even more money,
- 00:05:12you know, generally for right into Open
- 00:05:15AI. Um, how confident it's a lot of
- 00:05:17money, tens of billions of dollars. How
- 00:05:19confident are you that Open AI is going
- 00:05:22to be a financially sustainable and
- 00:05:24profitable enough company to justify all
- 00:05:26of this investment? It's looking like
- 00:05:29we're doing really well. I mean, like we
- 00:05:31have to we are definitely doing
- 00:05:33something unprecedented. Uh, but
- 00:05:36you know, it seems like a I like I feel
- 00:05:40confident in the bet. Doesn't mean
- 00:05:42something can't go wrong. You You
- 00:05:44tweeted that the GPUs were melting. You
- 00:05:48Yeah, I didn't realize people were going
- 00:05:48to take that literally. I I I mean, I
- 00:05:51get that it was a joke, but it's very
- 00:05:53hot, but like the metal is not actually
- 00:05:55melting. GPU shortage, bro. Um, that the
- 00:05:59team can't sleep. You're so busy. What
- 00:06:01happens when a launch goes viral and how
- 00:06:04does Stargate solve for this? Well,
- 00:06:08first of all, this level of viral this
- 00:06:09is an unusual thing. This last week, I
- 00:06:11don't think this has happened in the
- 00:06:12history of tech to any company before.
- 00:06:14Like the I don't know of any I've seen
- 00:06:18viral moments, but I have never seen
- 00:06:19anyone have to deal with an influx of
- 00:06:21usage like this. You added a million
- 00:06:22users in an hour or something like that.
- 00:06:24Yeah, I mean more some hours, but yeah.
- 00:06:27um it was like a it was like un
- 00:06:29unprecedented wild and also like making
- 00:06:32an image is not a it's not exactly like
- 00:06:34a low compute task the way we do it with
- 00:06:36the new image then so um we had to do a
- 00:06:39lot of very unnatural things we had to
- 00:06:40you know borrow compute capacity from
- 00:06:42research we had to slow down some other
- 00:06:43features because it's not like we have
- 00:06:45like hundreds of thousands of GPUs
- 00:06:47sitting around just like spinning idly
- 00:06:49so with if we had more GPUs we would be
- 00:06:53more able to
- 00:06:57handle demand surges like this and also
- 00:06:59we wouldn't have to put such
- 00:07:00restrictions on. Um I was just this
- 00:07:03morning making the list of what we have
- 00:07:05what we'd like to launch in the next few
- 00:07:08months and I was like you know I had the
- 00:07:11list over here and I had like my rough
- 00:07:12math of when we're getting GPUs here and
- 00:07:14where I thought we could get the
- 00:07:15efficiency gains and I was like this is
- 00:07:16not going to work like we're going to
- 00:07:19have to like pull some of these things
- 00:07:20make some trade-offs limit some things I
- 00:07:22don't like but this is not all of this
- 00:07:23is not going to fit here. So it's that
- 00:07:26directly linked. Yeah. I mean you can
- 00:07:28find efficiency gains and we tried to
- 00:07:30do. You can limit features and usage and
- 00:07:31we do that too. But uh you know more
- 00:07:33compute means we can give you more AI
- 00:07:35and you can use it for images or writing
- 00:07:37software whatever you want. Elon's got
- 00:07:39his own data centers and XAI just bought
- 00:07:42X which means he's got access to all of
- 00:07:44that additional data. He claims Grock is
- 00:07:47the smartest AI on Earth. How do you
- 00:07:49think it compares? I don't use it much.
- 00:07:52I think it's probably a very good model.
- 00:07:54Mhm. I mean obviously there are lots of
- 00:07:55competitors out there. There's a lot of
- 00:07:57good models. Uh
- 00:08:03I think I think good models
- 00:08:06will become very bountiful, very
- 00:08:07plentiful. Um so what's your edge going
- 00:08:09to be? The best infrastructure layer and
- 00:08:12the best top of the stack. You know we
- 00:08:14have way more people use chatbt than use
- 00:08:17any other uh AI service. Way more. Um,
- 00:08:21and I think you'll see with a lot of the
- 00:08:23new features we roll out that that will
- 00:08:25become um, more of an advantage over
- 00:08:27time. So once this is all built, what's
- 00:08:29the grand vision?
- 00:08:32Give people tools to let them do
- 00:08:34whatever they're going to do better. Um,
- 00:08:36you know, it's it's this is in some
- 00:08:37sense this is like it is different this
- 00:08:39time, but in many other senses, this is
- 00:08:41just like another tool and another piece
- 00:08:43of technological history. people will
- 00:08:46use it and unleash their creative energy
- 00:08:48and make all sorts of stuff that you and
- 00:08:50I love or some stuff that we don't. Um
- 00:08:52or stuff to just entertain themselves.
- 00:08:56Personally, the area I'm most excited
- 00:08:58about is AI for scientific discovery. Um
- 00:09:00I think that will, you know, we're not
- 00:09:02there yet, but we're not far away. I
- 00:09:04think 2025 will be a world where we have
- 00:09:07agents do a lot of work, but work of the
- 00:09:10kind of work and things we already know
- 00:09:11how to do. I'm hopeful that 2026 will be
- 00:09:13a big year of like really uh new
- 00:09:17scientific progress. Do you envision
- 00:09:19multiple Stargates on every continent?
- 00:09:21Like where else are we going to see you
- 00:09:23break ground? You will see them on other
- 00:09:24continents. Yes. Everywhere. I don't
- 00:09:27know about everywhere. Like this is this
- 00:09:29is like you you went to visit one of
- 00:09:31this like these are hard things to do.
- 00:09:33Well, look, this is an audacious bet on
- 00:09:35the future. What are the risks?
- 00:09:38Um
- 00:09:40I mean maybe you know people stop
- 00:09:42wanting to like pay for AI services and
- 00:09:44then we have a difficult financial
- 00:09:45position.
- 00:09:48How do you feel about one chipmaker
- 00:09:50Nvidia having so much power over the
- 00:09:52industry the future of the industry they
- 00:09:54make an incredible product and if you
- 00:09:57make an incredible product and people
- 00:09:58want it this is what happens to you. I
- 00:10:01do want to talk about the jobs thing
- 00:10:02because obviously we saw so many people
- 00:10:03working there. There is this lofty
- 00:10:05promise that AI data centers are going
- 00:10:07to create thousands and thousands of
- 00:10:08jobs. Meanwhile, AI is destroying jobs
- 00:10:11elsewhere. And I feel like even you know
- 00:10:13there is serious anxiety out there.
- 00:10:16People are scared. Totally. Even among
- 00:10:18the best engineers and technologists,
- 00:10:20people are scared. Um what do you say to
- 00:10:23those people?
- 00:10:24AI is for sure going to change a lot of
- 00:10:28jobs. Totally take some jobs away.
- 00:10:30Create a bunch of new ones. Mhm. This is
- 00:10:32like the kind of this is what happens
- 00:10:34with technology. In fact, if I think if
- 00:10:36you look at the history of
- 00:10:39the world like
- 00:10:41technological driven job change or
- 00:10:44whatever you call it when like one class
- 00:10:46of jobs go away and another one pop up
- 00:10:47like that's very consistent. It happens
- 00:10:49at a it's punctuated but like that's
- 00:10:51just been happening for a long time. And
- 00:10:54the thing that is different this time is
- 00:10:55just the rate with which it looks like
- 00:10:56it will happen. Mhm. Um, the thing I
- 00:10:59think the world is not ready for, like
- 00:11:00people have maybe abstractly thought
- 00:11:02like, okay, it's going to be a better
- 00:11:03programmer than me. It's going to be,
- 00:11:04you know, better at customer support and
- 00:11:06whatever. Um, I don't think the world
- 00:11:08has really had the humanoid robots
- 00:11:10moment yet. And I don't think that's
- 00:11:12very far away from like a visceral like,
- 00:11:14oh man, this is going to do a lot of
- 00:11:16things that people used to do. So, so
- 00:11:17yeah, it's coming. And we we have always
- 00:11:19tried to just be super honest about what
- 00:11:21we think the impact may be, realizing
- 00:11:23that we'll be wrong on a lot of details.
- 00:11:25What happens when the humanoid robots
- 00:11:26get here?
- 00:11:28Um,
- 00:11:31I mean they'll obviously do a lot of
- 00:11:32jobs, but what I the point I was trying
- 00:11:33to make is I think like the first day
- 00:11:35you're like walking down the street and
- 00:11:37there's like seven robots that walk past
- 00:11:38you doing things or whatever, it's going
- 00:11:40to feel very sci-fi. Mhm. Deepseek
- 00:11:43appears to have found a more efficient
- 00:11:44way to power AI. Was that a moment of
- 00:11:47rethink for you? And are you doing
- 00:11:49anything differently now? Um, I think
- 00:11:52the Deep Seek team is very talented and
- 00:11:53did a lot of good things. I don't think
- 00:11:55they figured out something like way more
- 00:11:56efficient than what we figured out. But
- 00:11:58do you think there is a more efficient
- 00:12:00way to build AI? I'm we will we have
- 00:12:03made incredible efficiency strides year
- 00:12:05over year and I'm sure we'll keep doing
- 00:12:07that in the future. So if that's the
- 00:12:08case, why are you building all this?
- 00:12:12If we had an AI that we could offer at
- 00:12:14onetenth of the price of current AI, um
- 00:12:17I think people would use it 20 times as
- 00:12:19much and we would still need twice as
- 00:12:21much compute to satisfy the then current
- 00:12:23demand. So let's talk about that Jevans
- 00:12:26paradox like how do you think Jevans
- 00:12:27paradox applies here that you know
- 00:12:30technological progress means
- 00:12:31paradoxically you're you're going to be
- 00:12:33using more resources doesn't
- 00:12:37it doesn't this never seemed like a
- 00:12:38paradox to me at all um like you know we
- 00:12:42talk about supply demand curves and
- 00:12:43elasticity and all sorts of other things
- 00:12:45and then people are like ah but I'm
- 00:12:46going to trick you with Jeb's paradox
- 00:12:47and it's like this is just the way the
- 00:12:49world works most of the time but you do
- 00:12:51think there will be more efficient ways
- 00:12:53what What are those? Is it better chips?
- 00:12:54Is it is it all of the above? It will we
- 00:12:57will have better chips. We will have
- 00:12:58better energy sources. We will have
- 00:12:59better algorithms. Um just we will
- 00:13:03optimize. I mean this is just this is
- 00:13:04like what industry does. We will
- 00:13:06optimize everything. Mhm. There's no
- 00:13:08question that China is going to be a
- 00:13:09formidable player in AI. How do you
- 00:13:12think about OpenAI's chances to win a
- 00:13:15global race?
- 00:13:19You know, probably I should have some
- 00:13:21deep thoughts about that.
- 00:13:23We're doing the best we can. Like we
- 00:13:25just wake up every day thinking how we
- 00:13:26can be a little bit better. Uh
- 00:13:28I don't know how to like I don't know
- 00:13:29how our actions would change based off
- 00:13:31of like some deep answer to that
- 00:13:33question. President Trump, you know, you
- 00:13:35said he cares about infrastructure.
- 00:13:36President Trump is in power at a pivotal
- 00:13:39time for AI development. What do you
- 00:13:41think his mark on this moment will be? I
- 00:13:43think he will get to make some of the
- 00:13:45most important decisions anyone in the
- 00:13:46world has gotten to make related to AI.
- 00:13:50Um, and you know, I'm optimistic he's
- 00:13:54really
- 00:13:55he'll really do the right thing there,
- 00:13:56but I I don't like he has an unenviable
- 00:13:59job, right? The velocity of in we were
- 00:14:03talking about the launches. The velocity
- 00:14:04of innovation here is just mind-blowing.
- 00:14:07The speed at which you're releasing new
- 00:14:09things, the speed of what's coming up,
- 00:14:11what I saw rising from the red dirt in
- 00:14:13Abolene. How do you personally grapple
- 00:14:16with the the pace of it all? That's one
- 00:14:19thing that's really impressed me about
- 00:14:20President Trump, by the way, is his
- 00:14:22ability to just like
- 00:14:26understand the whole industry and all
- 00:14:28the changes and quickly seem to have
- 00:14:30very good intuition and make good
- 00:14:31decisions about it. Uh, while things are
- 00:14:34changing so fast has has really been
- 00:14:36quite impressive to me. Um,
- 00:14:39but what about you personally? Yeah. I
- 00:14:40was going to say I
- 00:14:44I'm in it every day and and so I don't
- 00:14:48like
- 00:14:50it's like watching your own kid grow
- 00:14:51like you day to day you just see every
- 00:14:53change and so it's like not as striking.
- 00:14:56Um it does feel like it's going very
- 00:14:58fast. It's certainly
- 00:15:03I certainly think if I could like
- 00:15:04transport myself back three years ago,
- 00:15:07it would seem like unimaginable
- 00:15:09progress.
- 00:15:10But, you know, daytoday you can kind of
- 00:15:12get used to anything. Well, since you
- 00:15:14mentioned kids, you just had a baby. It
- 00:15:16did. Has that reframed anything for you?
- 00:15:20T. I mean, like, look, I don't think I
- 00:15:21have anything nonclée to say here, but
- 00:15:23it is the best most amazing thing ever
- 00:15:24and it totally rewired all of my
- 00:15:26priorities. I remember in like the first
- 00:15:28hour I like felt this neurochemical
- 00:15:30change and it happened so fast. I was
- 00:15:31like, "Oh, I get to like observe this.
- 00:15:33Like I am being like neurochemically
- 00:15:35hacked, but I'm like noticing it
- 00:15:37happening. I'm totally fine with this.
- 00:15:38This is working. is great, but like
- 00:15:40everything is going to be different now.
- 00:15:41What about how everything that you're
- 00:15:43doing here and building all this what it
- 00:15:46means for humanity? Has it reframed any
- 00:15:48of that? Um,
- 00:15:53a lot of people have said, "I'm very
- 00:15:54happy you're having a kid because I
- 00:15:56think you'll make better I think you'll
- 00:15:58make like better decisions to whatever
- 00:15:59degree you have." Like a you got to like
- 00:16:01make decisions here for humanity as a
- 00:16:03whole.
- 00:16:05I really wanted to get it right before
- 00:16:07I'd do the best I could. I I still
- 00:16:09really want to now. It uh
- 00:16:16it somehow Yes, it somehow does feel
- 00:16:18like it's different, but I can't
- 00:16:19articulate how. As you keep moving
- 00:16:22forward, if Chat GPT were to discover a
- 00:16:25Stargate, a futuristic wormhole that we
- 00:16:28could all travel through,
- 00:16:30what's on the other side? Like, what do
- 00:16:32you see that the rest of us don't?
- 00:16:35If Chip could like transport us to the
- 00:16:37future. Yeah. Through a Stargate. A
- 00:16:39wormhole. Please humor me on the
- 00:16:40wormhole. I have no idea. I mean, no,
- 00:16:43no, no one knows, right? Like, it's
- 00:16:44going to
- 00:16:48it's going to discover like AI. I I
- 00:16:50believe AI will help dramatically, like
- 00:16:53unbelievably dramatically accelerate
- 00:16:57the pace of science and human discovery
- 00:16:59and our understanding of what's
- 00:17:00possible.
- 00:17:02But like, do I think I could have sat
- 00:17:04here in 1905 and told you what we're
- 00:17:08about to discover in physics
- 00:17:10and that 40 years later we would like
- 00:17:12have an atomic bomb? Definitely not. Uh
- 00:17:16I and I I think I am way too self-aware
- 00:17:19of my own limitations to sit here and
- 00:17:21try to say I can like tell you what's on
- 00:17:23the other side of that wormhole. I have
- 00:17:24no idea. But net good. Yes. happy uh
- 00:17:28better a better a better world up and
- 00:17:31down not better in every way but yeah I
- 00:17:33think up to the right yeah up to the
- 00:17:35right with some choppiness Yes.
- AI
- Stargate
- 計算能力
- OpenAI
- 雇用
- 科学的発見
- 技術の進化
- Deepseek
- 資金調達
- 競争