large language model
also referred to as: large language models · llm
67 statements across 52 episodes · 30 bullish · 16 bearish · 50 people on the record · first statement Oct 27, 2023 by Artem Keydunov · said 1 times in 1 episodes since 2024 · across every show →
Mentions by year
brought up most by Jeremy Howard (1)
tap a year for its mentions
2024 1 mention in 1 episode
Everything said about large language model, oldest first
Oct 27, 2023 positive
Keydunov: Querying a semantic layer over raw SQL reduces LLM errors
“Now, the query, I believe, should be created against semantic layer, because it reduces the room for the error, because what usually happens is that your query to semantic layer would be very simple. It would be like, give me that metric grouped by that dimens…”
Dec 17, 2023 bullish
Yegge: AI coding will fragment into many specialized, fine-tuned models
“And that, that fragmentation of models actually, we expected to continue and proliferate, right? Because we are fundamentally, we're a recommender engine right now. We're recommending code to the LLM. We're saying, may I interest you in this code right here so…”
Dec 17, 2023 bullish
Liu: Reliable AI coding workflows require search-based algorithmic backbones
“The way that we get to this, like, more reliable, multi-step workflows that can do things beyond, you know, generate unit test is, is, it's really gonna be, like, a search-based approach, where, where you use an LLM as, kind of, like, an advisor or a proposal …”
Apr 27, 2024 negative
Bach: LLM consciousness simulations lack the functional properties of human brains
“I don't think that the consciousness in the LLM or the consciousness simulation in the LLM is has the same properties that it has in our brain. It does not have the same functional role, and it's also not implemented in the right way.”
Apr 27, 2024 positive
Apr 27, 2024 positive
Apr 27, 2024 neutral
May 31, 2024 negative
May 31, 2024 bearish
Huang: Adding one billion tokens cannot teach trillion-token models new knowledge
“All models these days are now double-digit trillions, right? So it's kind of a drop in the bucket if you really think I can just put, you know, a billion tokens in there, and I actually think that the model's gonna truly learn new Information.”
Jun 11, 2024 negative
Conover: LLMs do not need anthropomorphic personas for high-quality reasoning
“Our experience has been that You can get really, really high quality reasoning from roughly an agentic system without needing to be too cute about it. You can describe the task and you know, within well-defined bounds you don't need to treat the LLM like a per…”
Jun 11, 2024 negative
Conover: Unbounded AI agents are useless compared to finite state machines
“Specifically, like, I don't think that unbounded agentic behaviors are useful and that instead a useful LLM system is more like a finite state machine where the behavior of the system is occupying one of many different behavioral regimes and making decisions a…”
Jul 23, 2024 positive
Aug 28, 2024 positive
Carlini: Imperfect LLMs remain useful because users already distrust internet content
“You can't trust these things blindly, but I feel like most people on the internet already understand that things on the internet you can't trust blindly. And so there's not like, this is not like a big mental shift you have to go through to understand that it …”
Sep 20, 2024 negative
Schulhoff: Researchers should pay for top models instead of engineering routing
“For the most part, designing these systems where you're kind of routing to different levels of intelligence is a really time-consuming and difficult task, and, like, it's probably worth it to just use the smart model
And pay for it at this point if you're look…”
Sep 28, 2024 positive
Sep 28, 2024
Yan: LLM evaluation criteria cannot be determined without inspecting real outputs
“What they're saying is that it is impossible to completely determine good evaluation criteria without actually looking at LLM outputs. So essentially the point is you have to look at the data before you overcome evaluation criteria.”
Oct 11, 2024 neutral
Goyal: Figma vector formats consume far more LLM tokens than HTML or JSX
“Vectors are very difficult because they're a data inefficient representation, so the vector format in something like Figma Is choose up like many, many, many, many, many more tokens than HTML and JSX. So it's a very difficult medium to just sort of throw into …”
Oct 18, 2024 bearish
Drew Houston: Large language models are a rapidly self-commoditizing, bad business
“Large language models are a pretty bad
Business from a, you know, you sort of take off your tech lens and just sort of business lens. Like there's sort of this weirdly self-commoditizing thing where, you know, models only have value if they're kind of on this…”
Oct 25, 2024 negative
Martin: Chatbot novelty relies on human trickery, not interesting models
“Interacting with a chatbot is sort of novel at first, but it's not interesting. Right. And it's like humans are what makes interacting with chatbots interesting. It's like, ha ha ha, I'm going to try to trick it. It's like, that's interesting. Spell strawberry…”
Nov 11, 2024
Polu: LLM workflow development requires a dozen examples to prevent overfitting
“I had the strong belief from my research time that you cannot create an LLM-based workflow on just one example. Basically, if you just have one example, you overfit. So as you develop your interaction, your orchestration around the LM, you need a dozen example…”
Nov 11, 2024 neutral
Nov 25, 2024 positive
Nov 28, 2024 neutral
Nov 29, 2024 negative
Dec 24, 2024 positive
Ben Allal: Small models can generate synthetic data by rephrasing web pages
“The interesting thing in this approach is that you can use a model that is small Because it doesn't, rewriting doesn't require knowledge. It's just rewriting a page into a different style. So the model doesn't need to have like knowledge that is like extensive…”
Dec 24, 2024 positive
Ben Allal: LLMs can be trained with entirely synthetic pipelines
“Today you can train an LLM with like an entirely synthetic pipeline. For example, you can use our Cosmopedia data sets and you can train a one B model on like a hundred and fifty billion tokens. Those are a hundred percent synthetic, and those are also of good…”
Dec 25, 2024 bullish
Jan 10, 2025 positive
Bryk: Smaller LLMs with search tools are vastly superior to massive memorization models
“Yes, I believe this is a much more optimal system to have a smaller LLM that's really just like an intelligence module. And it makes a call to a search tool. That's way more efficient because if, okay, I mean, the opposite of that would be like the LLM is so b…”
Jan 26, 2025
Feb 1, 2025 neutral
Nguyen: Full Document Rewrites Yield Higher Model Accuracy Than Code Diffs
“We didn't know that, like, code diffs was very difficult for a model, for example. Again, it's like, do we go back to, like, fundamentally improve, like, code diffs as a model capability? Or do you, like, do a workaround where the model will just, like, rewrit…”
Mar 13, 2025
Shankar: Grounded synthetic data generation beats slow human annotation
“People don't know how to do anything other than Plan A, which is to go out and try to collect as much real world data as possible and take months because we're going to employ, like, human annotator teams to do this. Or Plan B, which is I'm going to ask an LLM…”
Mar 14, 2025
Ben-Smith: Snipd relies on regexes to format streaming LLM responses
“For this specific feature, like, we actually also have, like, countless regexes. That, that, they're just there to correct certain things that the LLM is doing, because it doesn't always adhere to the format correctly, and then it looks super ugly on the front…”
Mar 28, 2025 positive
Shah: Deterministic steps beat LLM randomness when workflow steps are known
“Thing number two is if you can get, if you know in your head what the actual steps are to accomplish whatever goal, why would you leave that to chance? There's no upside. There's literally no upside. Just tell me like, what steps do you need executed?”
Apr 23, 2025 positive
Build deterministic if-else LLM workflows instead of relying on prompt tuning
“I would emphasize people to focus on, like, building the most if-else condition-esque, like, LLM flows, because that gives you something to actually improve, i.e., like, if I identify an intent that's doing really shittily, but has high volume usage, I know th…”
Apr 24, 2025 bullish
Mlejnsky: LLMs will soon configure and provision their own cloud sandboxes
“And the goal where we think this is getting going is the LLM like decides what it wants to do and how it wants to have the sandbox configured. So it's basically starts controlling the infrastructure itself and creating sandboxes themselves.”
May 6, 2025 neutral
Kyutai's Bidirectional Moshi Architecture Has Not Scaled to Large LLMs
“The, that Kyutai Moshi model you mentioned, which was my favorite academic paper last year, is a step towards a truly bi-directional streaming in both directions, thinking all the time, LLM. That work has not been kind of scaled up to, you know, large LLM size…”
May 23, 2025 neutral
May 29, 2025 bullish
Jun 6, 2025 neutral
Ameisen: LLMs Execute Parallel Sub-Processes During Math and Hallucinations
“So I think one example of this is like math where the model is like independently computing the like last digit and then the like order of magnitude and then kind of like combining them at the end or like hallucinations are also that where like, there's one si…”
Jul 28, 2025 bullish
Mohan: Marginal loss improvements unlock massive gains in model reasoning
“Like, small wins at the margins are massive wins in terms of IQ. Like, it's harder to get those, and they don't look as big, but they're, like, massive wins in terms of reasoning. They can now do chain of thought, all these other things.”
Aug 5, 2025 negative
Dax Reed: Optimizing marginal LLM efficiency yields plateauing returns.
“I think the clarity we have is we're not trying to get the next three percent of efficiency out of the LLM. I think that's kind of like an infinite rabbit hole you can dive down that has like plateauing results. So for us, we're more focused on Product experie…”
Sep 1, 2025 positive
Sep 11, 2025 neutral
Martin: Context caching solves cost and latency, but not context rot
“I do think an important and subtle point here is that caching doesn't solve the long context problem. So it, of course, solves the problem of, like, latency and cost, but if you still have a 100,000 tokens in context whether it's cached or not, the LM is utili…”
Sep 11, 2025 positive
Martin: Tool-based search with llms.txt beats maintaining vector indexes
“You give an LLM access to simple files, file tools. In this case, I actually use an LLM.txt to help it out. So it can actually know what's in each file. It's extremely effective and much more simple and easy to maintain, easier to maintain than building an ind…”
Oct 2, 2025 bearish
Oct 2, 2025 positive
Oct 5, 2025
Oct 30, 2025 neutral
Oct 30, 2025 bullish
Sands: LLMs will cut Stripe payment integrations to two days
“The LPM team, it took them two weeks for the first one, but they just launched a new pan-European payment method, which In two weeks, using an LLM to, like, build that integration, and I think they'll probably, you know, have it down to a day or two within, wi…”
Nov 25, 2025
Li: LLMs can predict physical trajectories without abstracting physical laws
“I wouldn't be surprised that given enough celestial movement data, an LLM would actually predict pretty accurate movement trajectories... I wouldn't be surprised, but F equals MA or, you know, action equals reaction. That's just a whole different abstraction l…”
Dec 11, 2025 negative
Dec 26, 2025 positive
Dec 30, 2025 bullish
Nair: Continual learning during deployment does not risk model capacity overload
“If you could learn enough about those million tokens that you're actually in deployment on I don't think you should need, like, I don't think there's a risk of overloading the capacity of your model, right? Because you can train on a trillion tokens, and it's …”
Dec 30, 2025 positive
Nair: Context integration, not model intelligence, bottlenecks useful automation
“A big thing that needs to happen is, like, it's not, it doesn't feel like intelligence of the models is the bottleneck. It's more like you just have products that bring the entire context of what someone wants to do into the product so that the LLM can, like, …”
Jan 28, 2026 positive
White: LLMs Filter Scientific Hypotheses as Well as Human Domain Experts
“Nowadays, I actually would argue that if you go to an LLM and you ask it to evaluate, you know, hypotheses, including some garbage ones, it will probably do as good of a job as an expert in the field and filtering them out. That's not always the case.”
Jan 28, 2026 neutral
White: LLMs do not meaningfully accelerate physical CBRN and nuclear threats
“The classical example of nuclear is like, it's a lot of centrifugation, a lot of ultra centrifugation, a lot of high pressure or high RPMs. And so it, It's just, you can maybe get smarter about how to set up, you know, the economy of scale to do that with an L…”
Feb 12, 2026 positive
Jeff Dean: LLM search funnels trillions of tokens down to 100 documents
“And I think an LLM based system is not going to be that dissimilar, right? You're going to tend to trillions of tokens, but you're going to want to identify, you know, what are the 30,000 ish documents that are with the, you know maybe Thirty million interesti…”
Mar 9, 2026
Shah: True LLM User Understanding Requires Four Memory Capabilities
“So, to make an LLM truly good at understanding a user, you have to handle four things. One is knowledge updates, so you have to invalidate, steal knowledge, and build on top of it, which is different from the storing vectors. You have to have some sort of temp…”
Mar 24, 2026 bearish
Kulik: LLMs consistently fail to generate a 22-atom ligand
“The thing I constantly do every time an LLM is updated is I just ask it, please design me a ligand that has, ah, 22 atoms. So the first time I've done that, there are many ligands out there that have 22 atoms, and then I say, I want it to bind to the metal wit…”
Apr 18, 2026 positive
Swix: YouTube's recommendation system is LLM-based using tokenized video codebooks
“The YouTube Rexxus is LLM-based, and they- Is it really? They obviously, yeah. That's cool. It, they re, they tokenize every video, and put it in a code book, and then they train a LLM on it, and then feed in your context, just like a regular LLM, and ask it t…”
May 7, 2026 positive
Pocock: LLMs already understand Domain-Driven Design vocabulary
“Because it's been going around so long, it's kind of in the latent space of these models already, is if you say, you don't need to invent this whole new set of terms, you can just bolt on this framework, this system that's very, very flexible, very composable,…”
Jun 6, 2026 positive
Jun 6, 2026 negative
Jul 8, 2026 negative
Aug 3, 2026 neutral
Modern speech models operate autoregressively by adding waveform tokens to LLM vocabularies
“Speech is autoregressive. You effectively I mean, this was even back with, like, the Orpheus architecture a year and a half ago. You just add a bunch of waveforms to the vocabulary so that the LLM can output tokens that represent those waveforms, and then you …”
Aug 3, 2026 negative
Aug 15, 2026