GPT-4, every mention
80 scenes, the whole family · ← back to GPT-4
tap a year for its mentions
every year anyone John Coogan 68Flo Crivello 4Sam Altman 3Jordi Hays 3Will Brown 2Sholto Douglas 2Packy McCormick 2Marc Andreessen 2Everett Randle 2Aravind Srinivas 2
Verbatim, from the transcripts: the passages where GPT-4 comes up
Astra Reactions, Jobs Print, Cybercab Launch | Diet TBPN
- ▶ 9:43 unnamed speaker I think that was GPT four.
Meta’s AI Comeback Moment, Claude Mythos | Diet TBPN
Daniel Gross’ AGI predictions, SpaceX IPO news, Trump takes control of US chip exports | Diet TBPN
- ▶ 1:26 Jordi Hays He says, I think we can all agree that GPT four completes many tasks at human level proficiency. 2 times in the scene
The Remarkable AGI Predictions of Daniel Gross
- ▶ 1:23 unnamed speaker He says, I think we can all agree that GPT-IV completes many tasks 2 times in the scene
- ▶ 21:05 unnamed speaker DG says, two hundred and fifty billion of India's GDP exports are essentially GPT for tokens.
- ▶ 34:20 unnamed speaker And, uh, GPT-IV equivalent inference costs have collapsed. 3 times in the scene
The Secret Economics Controlling OpenAI, Anthropic, and the AI Boom
- ▶ 4:13 Packy McCormick I don't even know if the API is live anymore, but GPT four is the really instructive, uh, is the really instructive example because I believe that model costs like a hundred million dollars to train, and it was really expensive at the… 2 times in the scene
Full Interview: Clawdbot’s Peter Steinberger Makes First Public Appearance Since Launch
- ▶ 5:05 Peter Steinberger In May already, and I, I tried, it was like the time that GPT-FOR-ONE was out.
Capital One Acquires Brex for $5B, Davos’s $500M business, Apple’s Siri rebuild | Diet TBPN
- ▶ 9:04 John Coogan I made in one shot with GPT four.
X’s $1M Sweepstakes, Ben Affleck’s viral take on AI, Elon sues OpenAI for $134B | Diet TBPN
- ▶ 25:11 Ben Affleck I think it's because, and yes, it'll get better, but it's going to be really expensive to get better, and a lot of people are like, fuck this, we want you at GBD-IV, because it turned out, like, the vast majority of people who use AI are…
The Elon vs OpenAI Lawsuit and ads in ChatGPT | Diet TBPN
- ▶ 16:37 unnamed speaker Food and drinks would be served, but the main entree was a hush-hush demo by OpenAI co-founder Sam Altman on a forthcoming release of ChatGPT powered by GPT-IV, an AI built on large language models.
Amazon Bets $10B on OpenAI | News Reaction
- ▶ 21:29 unnamed speaker It's not like GPT-IV just came out.
The World’s Fastest Growing Defense Company, OpenAI’s Code Red, Google Strikes Back | Diet TBPN
- ▶ 12:33 Jordi Hays It's certainly, uh, people have, have, uh, uh, been very loyal to four.
- ▶ 21:15 John Coogan This is notable because OpenAI has had difficulty creating the next generation of models beyond the GPT-IV level of size and complexity.
Google Gemini 3 Reactions, Google Antigravity, Anthropic-Nvidia-Microsoft Deal | Diet TBPN
- ▶ 7:05 unnamed speaker The last time we saw a capability jump of this magnitude was the release of GPT-IV in March, 20, 23.
Blue Origin Lands Rocket, Valve Drops Steam Machine, Cantor Group Surges | Diet TBPN
- ▶ 26:35 John Coogan Just chat with four.
Thursday's Diet TBPN
- ▶ 15:32 unnamed speaker When GPT four, like, GPT four, like, not four or anything, when that was the best model, they would do updates, and they wouldn't, like, say, oh, this is a new model. 2 times in the scene
Sam Altman on Sora, Social Media, The Future of AI, and More
- ▶ 0:03 Sam Altman We know how to go make the GPT-IV equivalent of video models, and we will do that.
- ▶ 5:58 Sam Altman It was really not until GPT-IV where these text models started providing real value for people, and we know how to go make the GPT-IV equivalent of video models, and we will do that, and then a lot of these things that are currently… 2 times in the scene
- ▶ 18:24 John Coogan You're, you're not gonna be GP four. 2 times in the scene
- ▶ 22:31 John Coogan What do you think about the fact that people feel, at least, I don't know if they actually can, but it feels like you can still clock GPT-V writing, you know, it's not this, it's that, the M-dash, like, will we still see these artifacts in…
- ▶ 28:00 John Coogan Uh, I remember in the GPT, 3.5 GPT four days, it was like, don't build a company that assumes model stagnation.
Weekly Recap: Apple Debuts iPhone 17, Einstein of Wall Street, YC Demo Day, Declining AI Adoption
- ▶ 54:25 unnamed speaker Apple intelligence, you know, through that API, um, just the ability to actually distill something that's like GPT four level and have it be free, not even an API call that you need, not even cheap on a per token basis, um, but actually… 3 times in the scene
Weekly Recap: Sam vs. Elon, Cluely's Crazy Ad, UFC x Paramount, Apple's AI Robot
- ▶ 1:46:50 Everett Randle So I think that the way that you want to delineate it is, like, there's a certain, like, what I always tell everyone is that, like, there hasn't been a ChatGPT query since GPT-IV that, like, my mom hasn't been able to ask and have it… 2 times in the scene
9 Mansions Every AI Researcher Should Own
- ▶ 23:35 unnamed speaker So that's still pre-GPT three, like, pre-GPT four.
Weekly Recap: Apple Sitting on $100B, Disney Enters The AI Race, Marc Andreessen, All About GPT-5
- ▶ 5:30 John Coogan I'd like to, you know, put my credit card down and I want to be able to use the GPT for API.
- ▶ 1:04:04 John Coogan So like you, there is no world where you could just spend four million dollars and get this level model without also have done, having done the GPT four run, the GPT 4.5 run, et cetera, right?
- ▶ 1:11:23 Marc Andreessen Uh, I talked to somebody senior who was there at the time and I, I asked them, you know, when, when, when could you have had ChatGPT with GPT four level, uh, output, um, if you had just gotten, you know, gone, gone flat out starting in…
- ▶ 2:02:17 John Coogan There's been so much rhetoric around the, like, you know, that step up from GPT three to GPT four was insane.
- ▶ 2:14:16 John Coogan There was not it because previously it was like GPT three was this big GPT four was this big and you could visualize tangibly.
Marc Andreessen on Why Apple Might Miss the Next Platform Shift
- ▶ 5:47 Marc Andreessen Uh, I talked to somebody senior who was there at the time and I, I asked them, you know, when, when, when could you have had chat GPT with GPT four level, uh, output, um, if you had just gotten, you know, gone, gone flat out starting in…
Microsoft’s Cloud Outpaces Its AI Growth
- ▶ 5:23 unnamed speaker Like if they're, if they're, if you go to Azure and you say, I'd like to, you know, put my credit card down and I want to be able to use the GPT for API.
Weekly Recap: CEO Affair at Coldplay Concert, Elon Musk's AI Girlfriend, Cognition Acquires Windsurf
- ▶ 2:24:11 Martin Shkreli GPT four, same question.
Weekly Recap: Grok 4 Launch, Texas Floods, Web Browser War, Top Signals, Meta Smart Glasses
- ▶ 1:34:26 unnamed speaker And so this was like one of the, this was one of the key breakthroughs in like GPT four, right?
ELON MUSK's New Grok 4 Is Here! - We Break it Down
- ▶ 19:36 unnamed speaker And so this was like one of the, this was one of the key breakthroughs in like GPT-IV, right, was mixture of experts. 2 times in the scene
Apple Considers New AI partnership
- ▶ 7:02 John Coogan But I mean, I, I think the, I think the fact of the matter is like, uh, two billion dollars is enough to do a GPT four level run to do a deep seek style run. 2 times in the scene
Weekly Recap: Elon's Starship Explodes, OpenAI's Defense Contract, Nvidia's Robots, We Test Cluely
- ▶ 1:22:20 John Coogan Um, before we go into more of the consumer side, talk to me about some of the, uh, the more like, uh, niche, like B to B use cases, because I remember when like GPT, 3.5 and we got four, four, uh, GPT four dropped, there were still like a…
- ▶ 2:00:46 John Coogan when they really broke through with DaVinci and then four, like it felt like this kind of like binary moment when a lot of people realized that this was usable for their daily life, even if it's just a Google search replacement or…
Weekly Recap: Apple WWDC, Andrew Huberman, YC Demo Day, Intern Phone Assembly, $14B Meta x Scale
- ▶ 15:49 John Coogan GPT for a level, you know, you ask it a question and it reads you like a basic response and not lose the other Siri functionality.
Weekly Recap - Elon Vs Trump, Ukraine's Drone Attack, Cluely Update & OpenAI CRO
- ▶ 1:31:05 Sholto Douglas Um, I think you have like deep seek V three and this kind of stuff like R one, um, which means that with that's like at least two rooms just to get to the scale of GPT four and GPT four was two years ago. 2 times in the scene
How to Win in the Age of AI and Autonomy | Shyam Sankar on TBPN
- ▶ 4:52 Shyam Sankar Like you can't get the original GPT four anymore, right?
Why Humor Is the True Test of AI Intelligence | Will Brown on TBPN
- ▶ 7:49 Will Brown And it only now works, even though the models are the same, like GPT four was a bigger model than a lot of the models that we are using now.
- ▶ 10:57 Will Brown Like, the day GPT-IV launched, Morgan Stanley had integrations, because we had been working on it, and these were, like, we had press releases for these, like, we are ready to go, and so there are certain things where, like, initiatives…
Perplexity Founder Explains What Comes Next - Aravind Srinivas on TBPN April 23rd
- ▶ 7:35 Aravind Srinivas In fact, most people using tracks between the world don't even know there's a model called O one or O three and don't even know what the difference is for GPT four.
- ▶ 25:53 Aravind Srinivas You want to build at the right moment where, like, kind of, like, how we built perplexity around the GPT 3.5 time, not GPT four time.
Google's AI Comeback in Their Own Words - Logan Kilpatrick
- ▶ 1:32 unnamed speaker GPT-III to GPT-IV.
The Most UNDERRATED Use Case in AI | Jeff Huber on TBPN April 8th
- ▶ 4:37 Jeff Huber Um, for all kinds of reasons, you know, privacy, security, continuity, cost, you know, uh, you can build your startup on GPT-IV and it's amazing.
What AI and LEGO Have in Common | Flo Crivello on TBPN April 7th
- ▶ 11:29 Flo Crivello Well, you know, or like distillations, you know, like GPT-FOR turbo or like GPT-FOR is so much smaller than like the original GPT-FOR, like GPT-FOR was like a 1.7 trillion 3 times in the scene
- ▶ 11:29 Flo Crivello Well, you know, or like distillations, you know, like GPT-FOR turbo or like GPT-FOR is so much smaller than like the original GPT-FOR, like GPT-FOR was like a 1.7 trillion
Scott Wu on Cognition, Devin, and the Evolution of AI Agents.
How Does GROK Compare? (Full Analysis)
- ▶ 4:37 John Coogan Uh, Grok, uh, Grok says GPT four use 25 K GPUs while Grok three use 20,000 GPUs. 3 times in the scene
- ▶ 37:39 John Coogan I want to know the progression of GPT one to two to three to four for the pre-training runs. 6 times in the scene
- ▶ 46:46 John Coogan Like GPT-III to GPT-IV, more parameters, more data. 5 times in the scene
The BEST Super Bowl Commercial (Saquon Barkley)
- ▶ 37:46 John Coogan Although very debatable at this point, the latest number I saw was like, they spent 1.5 billion, which is like more than it costs to train GPT four.
The Future of Artificial Intelligence
- ▶ 6:36 unnamed speaker Yeah, I mean, like the, like, mixture of experts was, was described in detail by George Hotz when GPT-IV dropped, uh, and he was saying that there were some hardware limitations to how large you could scale the models, and so they were…
- ▶ 23:12 unnamed speaker If you want a GPT-IV model and you're training on all GPT-IV data or a lot of GPT-IV data, you're going to get a lot closer to GPT-IV pretty quickly. 3 times in the scene
DeepSeek Update, Market Crash, Timeline in Turmoil, Is VC Cooked, Zero Cope Policy
- ▶ 3:47 John Coogan So if you remember when GPT-IV was being rumored to drop, there were these, there was this massive, like viral meme image of like, here's a visualization of GPT-III, and it was like a small circle. 4 times in the scene
- ▶ 9:17 John Coogan So you see this massive GPT for, five hundred million dollar training run, huge data center, tons of networked and video chips altogether. 2 times in the scene
- ▶ 38:05 John Coogan It's very different when you're like, look, whoever gets to GPT four level models first, we'll have market entry into this new chat bot era, which is exactly what happened.
- ▶ 49:52 John Coogan And so basically, you know, for years, that's covered and working on ways to allow language models, like GPT four to solve tasks that involved reasoning, like math and science problems in 20, 21, he launched a project called GPT zero.
- ▶ 54:19 John Coogan Like if, if, uh, if deep seek trained on GPT four output tokens, then all the data has already been kind of cleaned because it's only training on like, of course it's going to sound like GPT four because it doesn't have any junk in there. 2 times in the scene
- ▶ 1:27:22 John Coogan You know, batch of, you know, tokens that it produces that you can't, I mean, it took him two years to pull all the data out of GPT-IV, right? 3 times in the scene
DeepSeek, Stargate Update, Bees Do Enjoy Honey, Golden Retriever Maxxing, Bring on the Trilly
- ▶ 4:23 John Coogan Today, many firms are within spitting distance of OpenAI's latest GPT-IV in chatbot ELO, and in some ways, and in some ways such as context length and video modalities, some firms are already ahead. 5 times in the scene
- ▶ 5:20 John Coogan Uh, so Gemini two ultra is rumored to surpass GPT four to turbo in every way.
- ▶ 19:46 John Coogan Um, OpenAI trained GPT-IV in twenty-twenty-two.
- ▶ 24:59 John Coogan This is more just like a GPT four competitor, which was built on a shoestring budget, which we'll discuss if that's real or not.
- ▶ 41:53 John Coogan I think the training data, a lot of it is synthetic data that's been pulled from the GPT-IV API.
- ▶ 1:19:04 John Coogan The, you can think about the GPT-IV cluster as five hundred million dollars.
Project Stargate, The T Word, Ramp Treasury, Audience Ad Reads, See You at Davos
- ▶ 19:33 John Coogan GPT-IV, which was a five hundred million dollar run, to a five billion dollar project, to the fifty billions coming, and the five hundred billion.
- ▶ 28:55 John Coogan It, you don't get the same increase from GPT three to GPT four.
- ▶ 35:01 John Coogan And that to my knowledge, no one has baked, you know, mid journey down to Silicon or, or GPT four down to Silicon. 2 times in the scene
The END of Ski Season? (E26)
- ▶ 2:19:08 John Coogan He ran the numbers on, for a while, when you got a response from GPT-IV, it would also give you some, like, not the full weights of the model, but like, a sampling of how that, like, response was calculated, and if you got enough of those,…
How we became the fastest growing podcast (ep20)
- ▶ 2:36:12 John Coogan I mean, the foundation model, yeah, it's still running GPT-IV, but obviously there's a lot of, uh, there's a lot of training to have that happens on top and a lot of UI stuff.
CEO Assassination Deep Dive (ep19)
- ▶ 55:31 unnamed speaker There's a New York Times article on the fact that, you know, GPT-IV can't tell you how many R's are in the word strawberry, and it gives you the real answer.
Inside Facebooks Secret "Red Book" (ep18)
- ▶ 1:52:02 John Coogan And, and so it's just a new way to, because they launched it with 3.5, which now if you use it, you're like, Whoa, this is way worse than four, but it was still magical. 3 times in the scene
- ▶ 2:01:28 John Coogan I mean, early on, I remember they, they were, uh, they were one of the huge beneficiaries of GPT-IV for the receipt scanning because they could just, just do exactly what we did with the red book where they, uh, would OCR it with Google,… 2 times in the scene
How BILLIONAIRES dress for XMAS (ep17)
- ▶ 3:03:02 John Coogan We looked at, we looked at like the, the ROI on GPT-IV trainings run.
Election Betting Gone WRONG (ep15)
- ▶ 39:41 unnamed speaker So he was very early in writing a story about Sydney being Sydney's, which was the first GPT four release.
Fact-Checking The Information (ep10)
- ▶ 1:12:35 John Coogan Yeah, and I bet most of the ChatGPT consumers that just go to chat.com, like, are blissfully unaware of the fact that there's a difference between four, four, oh, oh, one preview, oh, one mini. 2 times in the scene