GPT-3, every mention
59 scenes, the whole family · ← back to GPT-3
tap a year for its mentions
every year anyone Vishal Misra 19Erik Torenberg 10Amjad Masad 7Vijay Pande 6June Park 6Guido Appenzeller 5Eugenia Kuyda 4Chris Schroeder 4Steph Smith 3Mira Murati 3
Verbatim, from the transcripts: the passages where GPT-3 comes up
How Open Source Became AI's Backbone | Inferact with a16z
- ▶ 6:39 Matt Bornstein It's sort of, it's sort of interesting from a startup standpoint, um, uh, you know, like, like you mentioned, sort of GPT three or like early chat GPT, um, those closed source solutions were starting to become critical to like a small…
Why Scale Will Not Solve AGI | Vishal Misra - The a16z Show
- ▶ 1:06 Vishal Misra When GPD three was first released, uh, I got early access to it, and I started playing with it, and I was trying to solve a problem related to querying a cricket database. 6 times in the scene
- ▶ 11:03 Vishal Misra And when GPT-III was released, I thought maybe I could use StatsGuru, use GPT-III to create a front end for StatsGuru. 6 times in the scene
The AI Opportunity that goes beyond Models
- ▶ 2:20 unnamed speaker At that point in time, of course, ChatGPT
AI Eats the World: Benedict Evans on the Next Platform Shift
- ▶ 51:08 Erik Torenberg I'm, I'm curious as you reflect, you know, maybe post, uh, you know, chat GPT in two, or GPT three rather, um, 2 times in the scene
The 2045 Superintelligence Timeline: Epoch AI’s Data-Driven Forecast
- ▶ 33:02 Epoch AI Researcher Economic models I have seen of what happens if you get this sort of full replacement, you can automate a job, um, are, you know, either show this sort of an extremely fast wild takeoff, or with a couple of, or, um, you know, you have some…
Amjad Masad & Adam D’Angelo: How Far Are We From AGI?
- ▶ 7:31 Amjad Masad In the, in the true pre-training scaling era, you know, GPT-II, three, 3.5, maybe up to four, um, it, it felt like you, you can just, uh, put more internet data in there and just, it just got better.
- ▶ 38:44 Adam D'Angelo Um, the, the way we got to it was we, in early, we started experimenting with using GPT-III to generate answers.
Seeing The Future from AI Companions to Personal Software
- ▶ 36:55 Eugenia Kuyda Um, GPT-III before, you know, API to, to partner up with them to be one of the first partners, uh, for GPT-III API to launch. 4 times in the scene
Marc Andreessen & Amjad Masad on “Good Enough” AI, AGI, and the End of Coding
- ▶ 42:15 Amjad Masad And, um, and so I, I, I would have just expected, like, when we went from GPT-II to III, it was clear it was getting a lot more human. 2 times in the scene
Reid Hoffman on AI, Consciousness, and the Future of Labor
- ▶ 18:24 Reid Hoffman Like this is one of the things that like when I look at, you know, GBD two, three, four or five, it's a progression of savants.
Is AI Slowing Down? Nathan Labenz Says We're Asking the Wrong Question
- ▶ 5:03 Nathan Labenz And if you look back to GPT three, you know, there's a huge leap.
- ▶ 7:04 Erik Torenberg And so the difference between GPT two and GPT three and the GPT three and GPT four. 2 times in the scene
Will LLMs Get Us To AGI?
- ▶ 14:01 Vishal Misra So anyway, then the pandemic hit, the world stopped, but in July of 2020, the first version of GPD three was released. 6 times in the scene
- ▶ 16:44 Erik Torenberg And, and so maybe reflecting back since, since the release of GPT-III, what has most surprised you about how LLMs have, have developed? 4 times in the scene
- ▶ 22:06 Vishal Misra Now the size of this matrix is, you know, if you just take just the old, uh, first generation GPT-III model, which had a context window of 2000 tokens and a vocabulary of, uh, 50,000 next tokens or 50,000 tokens, then the size of it, the…
Sam Altman on Sora, Energy, and Building an AI Empire
- ▶ 32:38 Erik Torenberg Um, cause there's been some evolution in thinking too, and that GPT-III didn't have the open, open weights, but you released a, you know, very capable open model earlier this year.
Building an AI Physicist: ChatGPT Co-Creator’s Next Venture
- ▶ 16:36 Anjney Midha There was a moment in time when models like, or papers like the GPT-III paper, for example, that
From Vibe Coding to Vibe Researching: OpenAI’s Mark Chen and Jakub Pachocki
- ▶ 1:28 Mark Chen You had, uh, the GPT kind of two, three, four series, which were kind of these instant response models.
- ▶ 3:00 Jakub Pachocki I think another thing that's maybe even more important, but a little bit subtler, when we were in this like GPT-II, GPT-III, GPT-IV era, um, you know, there was kind of one recipe you just like
- ▶ 37:19 Jakub Pachocki if you, if you look at GPT-Tree, right?
Dylan Patel on the AI Chip Race - NVIDIA, Intel & the US Government vs. China
- ▶ 1:16:49 Dylan Patel And so like, you know, it was like really impressive when GPT, uh, you know, two was trained on so many chips and then GPT three was trained on that, you know, like on, on, on 20 K 100
Faster Science, Better Drugs
- ▶ 15:28 Patrick Hsu I find it helpful to frame these in terms of like GPT one, two, three, four, five capabilities, right? 2 times in the scene
- ▶ 25:32 Erik Torenberg I want to hear your answers to a couple of these questions too, as, as lead of our bio practice, both on the GPT-III moment, what that, what that could look like, and also, um,
How Scale AI is Pioneering the Future of Work
- ▶ 45:35 Ben Scharfstein GPT three drafts off of, you know, GPT two and four, whatever, five, but it's like going to be pretty irrelevant later.
Aaron Levie and Steven Sinofsky on the AI-Worker Future
- ▶ 50:51 Aaron Levie Um, and I, I think it was, I think it was basically, uh, a phenomenon in the first wave, which was if you were just doing, like, basic, like, like, if you had figured out that you could do something on GPT, you know, two and three, where…
The Current Reality of American AI Policy: From ‘Pause AI’ to ‘Build’
- ▶ 16:11 Martin Casado GPT-II comes out, and then III comes out, and like, oh, this stuff's amazing, and somehow it got conflated.
GPT-5 and Agents Breakdown – w/ OpenAI Researchers Isa Fulford & Christina Kim
- ▶ 19:37 Sarah Wang Do you think the jump from GPT four to five was bigger, or three to four, or maybe three and a half to four?
- ▶ 35:01 Christina Kim Honestly, I kind of had this moment before I joined OpenAI, like, like, I think with the scaling laws paper with GPT-III, I was just like, kind of hit me that, like, if this exponential is true, like, there's not really much else I want to… 3 times in the scene
Dwarkesh Patel and Noah Smith on AGI and the Economy
- ▶ 46:39 Dwarkesh Patel GPT three can technically do a lot of things.
Tech vs. Media: Balaji Srinivasan on the Battle Shaping Our Future
- ▶ 1:13:30 Balaji Srinivasan And I actually feel pretty good about that, essentially predicting that GPT-three would be able to summarize, the next version of it would be able to summarize things.
- ▶ 1:21:28 Balaji Srinivasan And the reason I knew GPT-III might get there is there's a company called Narrative Science, actually.
Balaji Srinivasan: How AI Will Change Politics, War, and Money
- ▶ 16:46 Balaji Srinivasan You know, one of the, one of the surprises I had, that's why it's interesting when you said economics, one of the surprises I had with GPT-III and ChatGPT was how far you could get with language.
Reasoning Models Are Remaking Professional Services
- ▶ 7:06 George Sivulka And it said GPT-III. 3 times in the scene
- ▶ 9:13 Alex Immerman You were talking about GBT-III came out in summer of twenty-twenty.
Marc & Ben on AI Policy, Safety, Censorship & Unexpected Risks
- ▶ 24:34 Ben Horowitz And if you look at like the improvement from GPT, uh, two point O to GPT three to 3.5, and then compare that from like 3.5 to four, you know, we've really slowed down in terms of the amount of improvement.
Why Human Data is Key to AI: Alexandr Wang from Scale AI
- ▶ 4:04 Alexandr Wang So phase one, you, you know, sort of hallmarks are the original transformer paper, um, the original small scale experiments on GPTs, um, you know, all the way leading up, probably until like GPT three, 3 times in the scene
7 Ways to Boost Retention (Both Pre- and Post-AI)
- ▶ 5:05 unnamed speaker And there's an example of an AI-native companionship product where they actually did this test where they used GPT-one, GPT-two, GPT-three, and they actually tracked the cohort of user retention using each of the models. 2 times in the scene
Bringing AI to the Masses with Adam D'Angelo, CEO of Quora
- ▶ 2:58 Adam D'Angelo We, we ran some experiments using GPT-III to generate answers and compare them to the, the answers that, that humans had written on Quora. 2 times in the scene
Safety in Numbers: Keeping AI Open
- ▶ 3:25 Arthur Mensch There was also a misconception on GPT-free and basically in 20, 21, every paper, uh, made this mistake.
- ▶ 14:34 Arthur Mensch And all of a sudden in, uh, with GPT-free, this tide kind of reversed and, uh, companies started to be more opaque about what they were doing because they,
Inside AI Town: What AI Can Teach Us About Being Human
Where We Go From Here with OpenAI's Mira Murati
- ▶ 12:01 Mira Murati You know, after we developed GPT-III and we put it out there in the API, this was the first time that we actually had safety research become practical into the real world, and this happened through instruction following models. 3 times in the scene
Improving AI with Anthropic's Dario Amodei
- ▶ 4:02 Anjney Midha Sort of famously what happened then was you, you saw a reason to continue down that line of inquiry, which resulted in GPT-III. 2 times in the scene
The True Cost of Compute
- ▶ 4:45 Guido Appenzeller So if you take something like GPT-III, right, which is, you know, the open air is big model. 2 times in the scene
- ▶ 6:04 unnamed speaker GPT-III. 4 times in the scene
Chasing Silicon: The Race for GPUs
- ▶ 16:37 Guido Appenzeller The, you know, the GPT-III was a hundred seventy-five billion parameters. 3 times in the scene
- ▶ 21:52 unnamed speaker We'll explore all this in depth, including how much startups are really spending on AI compute and whether that's sustainable, how much it really costs to train a model like GPT-III, the difference in cost between training and inference,
Global Investing in Tech: Pattern Recognition, Founder Traits, Momentum | Field Notes by Connie Chan
- ▶ 25:34 Chris Schroeder And the second was GPT-III. 4 times in the scene
The Robot Lawyer Resistance with Joshua Browder of DoNotPay
- ▶ 26:13 Joshua Browder These generic AI models like, um, DaVinci from OpenAI, they're, they're not actually that good at the law.
The 1000x Developer with Amjad Masad
- ▶ 21:12 Steph Smith So something like a co-pilot is built on top of GPT-III to my knowledge, and that's a decision to be built off another platform, but you went a different route.
- ▶ 33:27 Amjad Masad Right now, you can just use GPT-III to do the parsing. 2 times in the scene
- ▶ 39:14 Amjad Masad GPT-III is incredibly good at parsing. 2 times in the scene
2023 Big Ideas in Technology (Part 1)
- ▶ 1:06:13 Steph Smith Because right now we are relying on a few models like GPT-III, which influence Copilot or some of these other layers. 2 times in the scene
- ▶ 1:14:17 Samit Singh While potential use cases within fintech are still emerging, LLMs like GPT-III and the upcoming GPT-IV may help businesses train datasets much more quickly and cheaply.
Expert AI as a Healthcare Superpower
- ▶ 5:48 Vijay Pande Like, can, people can take, like, GPT-III, you can give it SAT exams, can do okay, can, like, ah, 6 times in the scene
The Factory of the Future with Chris Power
- ▶ 22:45 Chris Power Like good luck running compute on GPT-III if there are no chips or China takes Taiwan and then the only chips you can buy have like a bug in them.
AI and the Creator Economy with Karen X Cheng
- ▶ 55:47 Karen X. Cheng Um, in fact, I was actually just talking to someone who was using GPT-III to write prompts to run through stable diffusion.