large language model

also referred to as: large language models · llm

56 statements across 48 episodes · 16 bullish · 21 bearish · 44 people on the record · first statement Sep 13, 2022 by Gary Marcus · across every show →

Everything said about large language model, oldest first

Sep 13, 2022 bearish
Prediction Held up
Marcus: A death will be tied to an LLM within a year
“The prediction that I made Is basically that there will be a death tied to a large language model in the next year.”
Gary Marcus Sep 13, 2022 ▶ 48:29 Is AI Dangerously Overhyped? — With Gary Marcus
Nov 9, 2022 positive
Opinion
Aguera y Arcas: LLMs legitimately understand concepts without having personhood
“They can understand concepts but you know, this is not, it's not personhood.”
Blaise Aguera y Arcas Nov 9, 2022 ▶ 3:11 Fantasy and Fact: AI, Sentience and The Dangers Of Hype, from World Summit AI
Feb 23, 2023 neutral
Prediction Not checkable as stated
Marcus: LLMs will connect to all major software within two years
“You have companies like Adept that are spending their whole time trying to quote, add all the world, connect all of the world's software to large language models. So if not now, soon enough. I think that that's right. Like, you know, two years if it's not toda…”
Gary Marcus Feb 23, 2023 ▶ 25:55 Blake Lemoine and Gary Marcus Debate AI Chatbots
Feb 23, 2023 negative
Insight
Marcus: LLMs hallucinate because they cannot distinguish individuals from categories
“The way I think about it is that these things don't understand the difference between individuals and kinds. So I actually wrote about this twenty-some years ago in my book, The Algebraic Mind, and I gave an example there, which is I said, suppose it was a dif…”
Gary Marcus Feb 23, 2023 ▶ 5:48 Blake Lemoine and Gary Marcus Debate AI Chatbots
Feb 23, 2023 neutral
Insight
Lemoine: Reinforcement learning transforms language models into goal-oriented systems
“The subsequent fine-tuning, and especially once you add reinforcement learning, it's no longer just trying to predict the next token in a stream of text. Specifically in the reinforcement learning paradigm, it's trying to accomplish a goal.”
Blake Lemoine Feb 23, 2023 ▶ 3:22 Blake Lemoine and Gary Marcus Debate AI Chatbots
Aug 3, 2023 positive
Opinion
Wood: Using LLMs for search ranking presents a larger commercial opportunity
“The next area, which is Less sexy, but in my opinion, maybe even be a larger opportunity is using this technology to improve search results, improve ranking, relevance, personalization, those sorts of use cases where you don't even know that you're working wit…”
Matt Wood Aug 3, 2023 ▶ 22:15 Amazon Reveals Its AI Master Plan — With Matt Wood
Aug 3, 2023 neutral
Opinion
Wood argues public web data is fair game for AI model training
“Who's to say what this data can be used for when it's publicly available. You know, you chose to make it publicly available. If you want to put some permissions around it, or you want to take it private, Yeah, that's totally up to you. You still own the data. …”
Matt Wood Aug 3, 2023 ▶ 42:40 Amazon Reveals Its AI Master Plan — With Matt Wood
Dec 12, 2023 positive
Insight
Grounding LLMs With Proprietary Data Reduces Hallucinations and Improves Caching
“Marrying the large language model with our own data sets with third-party data really drives down, ah, the amount of hallucinations that we get from these kinds of conversation and almost creates authoritative sources of cash information around hotel reviews, …”
Marty Brodbeck Dec 12, 2023 ▶ 10:27 How Generative AI Is Changing Travel — With Priceline's Marty Brodbeck
Dec 12, 2023 negative
Insight
Unmanaged LLM API Calls Cause Generative AI Costs to Skyrocket, Brodbeck Warns
“Cost in, in running and managing a generative AI platform could skyrocket. So you have to be very smart about when you refresh or make direct calls to the large language model and this third party information.”
Marty Brodbeck Dec 12, 2023 ▶ 10:59 How Generative AI Is Changing Travel — With Priceline's Marty Brodbeck
Mar 20, 2024 neutral
Insight
Doshi: Computer vision lacks a single unified model equivalent to LLMs
“This is missing in vision, but definitely kind of exists in language and language. We can solve hundreds or thousands of different tasks. But in graphics, but in vision, it's all separated. It's kind of like where language was back three or four years ago wher…”
Suhail Doshi Mar 20, 2024 ▶ 19:53 Predicting AI’s Next Advances — With Suhail Doshi
May 1, 2024 neutral
Insight
AWS VP: Pure LLMs Won't Deliver Effective AI Reasoning Without Other Tools
“Reasoning capability is not, if we view it as purely like everything is going to be LLM driven, we as industry are going to be very, very not satisfied. That's why it has to be actually more iterative. And we work with LLM as one of the ingredients. It is not …”
Swami Sivasubramanian May 1, 2024 ▶ 29:39 AWS VP of AI and Data Swami Sivasubramanian — GenAI's Growth Potential
May 15, 2024 bearish
Insight
We cannot be certain current AI development is actually tracking to AGI
“You shouldn't be, like, sure that they're gonna we're on the track of AGI because, yeah, fundamentally we don't know what kinds of things these are.”
Dwarkesh Patel May 15, 2024 ▶ 30:44 AI Scaling, Alignment, and the Path to Superintelligence — With Dwarkesh Patel
May 22, 2024 bullish
Insight
Aaron Levie: LLMs serve as the common reasoning engine across agent architectures
“If you go and ask like 10 AI startups that are doing anything with agents, you'll probably get, you know, more or less 10 different architectures of kind of how the agent actually functions. But what is at least similar to all of them is the LLM or the model i…”
Aaron Levie May 22, 2024 ▶ 18:18 Box CEO Aaron Levie — What Cheaper, Faster, and Smarter AI Gets Us
Jul 17, 2024 negative
Insight
Siemiatkowski: LLMs regress toward the average and fail at breakthrough marketing copy
“If I want some amazing copy that, you know, articulates what Klana does that's different, I would not rely on ChatGP to do that. I would rely on a human because it, a human has a bigger, is much better at kind of being far-fetched and do something crazy and ou…”
Sebastian Siemiatkowski Jul 17, 2024 ▶ 31:23 Did Klarna Really Automate 700 Jobs With AI? — With Sebastian Siemiatkowski
Jul 24, 2024 neutral
Prediction Not checkable as stated
Future software will combine multi-modal AI models with traditional code
“We're going to see computer systems where large language models are just one building block in addition to code, or maybe it's multiple language models and image models and, you know, time series models and whatnot, plus code combined. To generate all the ou…”
Thomas Domke Jul 24, 2024 ▶ 9:14 GitHub CEO Thomas Domke -- The One-Person, Billion-Dollar Startup
Jul 31, 2024
Disclosure
Shapiro: NVIDIA uses LLMs to turn accident reports into AV simulations
“The other thing that we can do is we can take accident reports and now using these large language models, we can input these accident reports and be able to create scenarios from a text input explaining what happened or if there's a map or something like that.”
Danny Shapiro Jul 31, 2024 ▶ 18:47 NVIDIA's Auto Play and the Future of Autonomous Driving — With Danny Shapiro
Sep 17, 2024 bearish
Prediction Not checkable as stated
Rhee: Standalone AI Models Won't Make Money at Expected Scale
“AI itself, will it make money itself, the model, and LLM? Some will, I believe, but not in the scale that we're looking at. Really, it's going to come down to where we apply the AI, and it doesn't matter which sector it is.”
Sokwoo Rhee Sep 17, 2024 ▶ 27:24 Sparking Net New Innovation At Massive Scale — With LG's Sokwoo Rhee
Oct 4, 2024 negative
Opinion
Bosworth: LLMs cannot be conscious and are merely associative databases
“And also I do not believe it is even the kind of thing from which consciousness exists. It doesn't exist in any Again, to my point about Doug Hofstetter it doesn't exist in any self-referential loop. In fact, it doesn't exist at all except when I instantiate i…”
Andrew Bosworth Oct 4, 2024 ▶ 31:42 Meta's Chief Technology Officer On Its Augmented Reality Dreams — With Andrew Bosworth
Oct 11, 2024 neutral
Insight
Kantrowitz: AI's High Compute Costs Break Traditional Software Accounting Models
“There's no standard accounting practice for how to amortize The value of a large language model because it's so different in the past. You build software, you have like essentially zero marginal cost. It can go forever. That's it. This is a much more cost inte…”
Alex Kantrowitz Oct 11, 2024 ▶ 15:26 OpenAI's Leaked Financial Data
Nov 27, 2024 neutral
Insight
Benioff: AI cannot run solely on LLMs without grounding in metadata
“As we get into this next level of AI, it needs the data to be intelligent. It can't just run off of a large language model, which is an intelligence model. It has to be what we call grounded or connected into the customer data itself. And not just the customer…”
Marc Benioff Nov 27, 2024 ▶ 16:37 Salesforce CEO Marc Benioff Has Thoughts on AI Agents, Automation, and The Future of Your Job
Jan 5, 2025 bullish
Prediction Not checkable as stated
Newton predicts at least one AI company will reap massive LLM profits
“So when I look at AI, I see something that yes, is very expensive to train, but when you look at the fact that it's a general purpose technology and that we are finding more and more things to do with it, I do believe that at least one of these companies is go…”
Casey Newton Jan 5, 2025 ▶ 3:12 Bold Predictions For Tech and AI in 2025 — With Casey Newton
Feb 27, 2025
Assertion Not checkable as stated
Chen: Nearly all large language models today utilize mixture of experts
“I think pretty much all large language models today use, utilize mixture of experts.”
Mark Chen Feb 27, 2025 ▶ 11:38 OpenAI's Chief Research Officer on GPT 4.5's Debut, Scaling Laws, And Teaching EQ to Models
Mar 5, 2025 neutral
Insight
Daniel Rausch: Alexa orchestrates multiple specialized models rather than one LLM
“People assume the large language model is the product. A product like Alexa is. So much more than quote unquote, just a large language model. So you have models playing many different roles in the system overall, even models helping us decide which model and m…”
Daniel Rausch Mar 5, 2025 ▶ 9:04 How Amazon Rebuilt Alexa From The Ground Up
Mar 14, 2025 negative
Insight
Kantrowitz: Apple's closed, deterministic product culture cannot adapt to open-ended AI
“It's an excellence that makes products work very well with their custom silicon and the fact that their operating system is closed and they control it. That's a great innovation and it works well with their operating system. But that type of excellence is not …”
Alex Kantrowitz Mar 14, 2025 ▶ 12:45 Gruber Slams Apple's Siri Failure, Tech Stocks In Freefall, Vibecoding & OpenAI's Agents
Mar 19, 2025 bearish
Insight
LeCun: Large language models are incapable of inventing new things
“We also know that they can hallucinate facts that aren't true but they're really, in their purest form, they are incapable of inventing new things.”
Yann LeCun Mar 19, 2025 ▶ 1:45 Why Can't AI Make Its Own Discoveries? — With Yann LeCun
Apr 7, 2025 positive
Assertion Supported
George: LLMs Dynamically Generate and Adapt IT Automation Scripts
“Instead, the LLM is automatically generating that execution plan so it can be run. And even better, as you run that script the LLM is also determining if you encounter Behavior that you didn't expect. It's resilient enough to be able to work around that.”
Joseph George Apr 7, 2025 ▶ 9:02 Understanding Practical AI and the Future of Automation – With Joseph George
Apr 16, 2025 neutral
Insight
Shao: Once LLMs become good enough, AI competition shifts to products
“If we've reached a point where everything is kind of just good enough, then I think commercialization is also going to be in product layer.”
Grace Shao Apr 16, 2025 ▶ 21:41 Why China's AI Strategy Just Might Work — With Grace Shao
May 14, 2025 bearish
Opinion
Bender says large language models are parlor tricks exploiting human perception
“Right down at the bottom, you've got the fact that especially large language models are a technology that is, that's a parlor trick. It plays on our ability to make sense of language and makes it very easy to believe there's a thinking entity inside of there.”
Emily M. Bender May 14, 2025 ▶ 1:05 AI’s Drawbacks: Environmental Damage, Bad Benchmarks, Outsourcing Thinking
May 24, 2025 neutral
Opinion
Kavukcuoglu: No major AI lab relies solely on scaling LLMs for AGI
“But also, I don't think that there's any research lab that is trying to only do scaling up the LLM.”
Koray Kavukcuoglu May 24, 2025 ▶ 9:45 Google DeepMind CTO: Advancing AI Frontier, New Reasoning Methods, Video Generation’s Potential
Jun 14, 2025 bearish
Prediction Not checkable as stated
Roy: Manual data labeling by Scale AI will no longer be relevant
“What they did, their job, again, like actually having people manually tag, like large networks of people manually tagging data to make it more ingestible for a large language model or for a training, It's not going to be as relevant anymore. You can even now h…”
Ranjan Roy Jun 14, 2025 ▶ 40:38 Sam Altman’s Gentle Singularity, Zuck’s AI Power Play, Burning Of The Waymos
Jul 1, 2025 negative
Prediction Not checkable as stated
Anugas: Enterprises will avoid public LLMs like ChatGPT for isolated data stores
“One of the things that's important that we really want everyone to see is a future where you're not sending your data to an LLM. You're not going to load your data into chat GPT. You're not going to do that. Okay. You're going to want that in an isolated data …”
Greg Anugas Jul 1, 2025 ▶ 33:02 An AI Brain For Your Business?
Jul 4, 2025 positive
Insight
Dialogic AI offers doctors a real-time interpretability mechanism into model reasoning
“The dialogic nature means that a human doctor can follow along and actually learn In a very transparent way. It's almost like having an interpretability mechanism inside the black box of the LLM, because you can see its thinking process in real time.”
Mustafa Suleyman Jul 4, 2025 ▶ 7:36 Microsoft AI CEO Mustafa Suleyman: Our AI Doctor Outperforms Human Diagnosticians
Aug 4, 2025 negative
Opinion
Kantrowitz: OpenAI and Meta are building the exact same thing using LLMs
“I'll stand on the table here and say it's the same thing. I think it's, if there were different techniques, I think this is really what it comes down to. If there were different techniques to build AGI and super intelligence, I would say, okay, sure, you know,…”
Alex Kantrowitz Aug 4, 2025 ▶ 28:02 Mark Zuckerberg’s Personal Superintelligence, Layoffs & Payoffs, Writing With AI — With M.G. Siegler
Sep 10, 2025 positive
Assertion Partly supported
LLMs reduced Veterans Affairs claims research time from hours to seconds
“We have large language models helping the VA do claims processing. What used to take many hours for a person researching on the claim process happens in a few seconds through The use of a large language model.”
Bill Vass Sep 10, 2025 ▶ 20:53 Booz Allen CTO: AI in Government, Autonomous Driving, Quantum's Promise
Sep 10, 2025 positive
Assertion Not checkable as stated
Amazon successfully used manual-referencing LLMs to debug fulfillment centers without hallucinations
“We use large language models at Amazon to help debug things in our fulfillment centers. And it was very successful in those areas, but you still have, it references the manuals directly so you can avoid hallucination. You can see what it actually, you know, fo…”
Bill Vass Sep 10, 2025 ▶ 27:17 Booz Allen CTO: AI in Government, Autonomous Driving, Quantum's Promise
Sep 18, 2025 neutral
Assertion Supported
Kantrowitz: MIT study shows 90% personal AI use vs 40% corporate purchases
“Official LLM purchases cover only 40% of firms, yet 90% of employees use personal AI daily, at least those surveyed”
Alex Kantrowitz Sep 18, 2025 ▶ 12:39 Are 95% of Businesses Really Getting No Return on AI Investment? — With Aaron Levie
Oct 14, 2025 negative
Insight
Harris: Enterprises cannot dump all data into LLMs without security models
“You can't just like, I want to take all the data of an enterprise and dump it in one place, you know, and you're not going to dump it into an LLM because LLM has no security model.”
Parker Harris Oct 14, 2025 ▶ 9:27 Will AI Kill Software? — With Salesforce Co-Founder Parker Harris
Oct 15, 2025 negative
Assertion Supported
Summers: Attackers Use Hidden White Text to Exfiltrate Enterprise LLM Data
“And just in the last few weeks, we've seen these really innovative Attacks where somebody will open up you know, something that's got instructions to an LLM. So it's processed within an LLM, but you know, it's hidden. It's like white text on a white background…”
Grady Summers Oct 15, 2025 ▶ 16:47 Is the AI Boom About to Break Security? — With Grady Summers, CEO of Netwrix
Oct 29, 2025 neutral
Insight
Kelleher: Autonomous action distinguishes AI agents from LLMs and brings security risks
“I think when we talk about the distinction between an artificial intelligence, the distinction between a large language model and an agent, I think the ability to take action is a very specific capability of an agent, and it's really the ability to take autono…”
Eric Kelleher Oct 29, 2025 ▶ 5:43 AI Agents: Hype or Hope
Oct 29, 2025 neutral
Insight
Kelleher: Key AI governance fear is data leaking across queries
“One of the latent concerns people have as they're deploying, deploying third-party agents, is they want to know if that data is going to be used anywhere else. So is it going to be used just for my query to help me do my work, which is usually fine? Or is that…”
Eric Kelleher Oct 29, 2025 ▶ 21:58 AI Agents: Hype or Hope
Oct 30, 2025
Assertion Not checkable as stated
Medium's Content Corpus Is Large Enough to Poison Any AI Model
“We learned really early on that medium is a big enough corpus of data that we can poison the the results of any large language model.”
Tony Stubblebine Oct 30, 2025 ▶ 16:15 Medium CEO Tony Stubblebine: Here's How Writing Evolves In The AI Era
Nov 5, 2025 bullish
Prediction Not checkable as stated
Sansbury: Threshold for Human Intervention in AI Workflows Will Keep Rising
“I will tell you if the improvement in the large language models that I use is an indication, the line's going to keep going up. The technology is iterating very fast and getting better much more quickly than I would have thought.”
Charles Sansbury Nov 5, 2025 ▶ 16:47 Why Getting Data Right Could Be The Key To Effective AI Projects — With Charles Sansbury
Nov 25, 2025 positive
Opinion
Multimodal AI Systems Are Better Described as 'World Models' Than LLMs
“So I think the very description of LLM Needs to change. In fact, I think now we're talking about world models. I think world models are a more accurate description of what we're talking about.”
Mark Moffat Nov 25, 2025 ▶ 15:31 Will AI Make Its Biggest Splash In Industrial Use Cases? — With Mark Moffat, IFS
Dec 3, 2025 negative
Insight
Hubinger: Training models to cheat triggers latent concepts about human badness
“When the model learns to cheat, when it learns to cheat on these programming tasks, it causes these other latent concepts about, you know, misalignment and badness of humans to bubble up, which is surprising, right? You might not have initially, you know, tho…”
Evan Hubinger Dec 3, 2025 ▶ 49:58 How An AI Model Learned To Be Bad — With Evan Hubinger And Monte MacDiarmid
Dec 5, 2025 neutral
Prediction Not checkable as stated
McDermott: True AI Memory May Require Tech Beyond Current LLMs
“We might not get there with the exact technology that we're using for large language models at the moment, but it feels like some way that we want to get to with sort of frontier level at AI.”
Adrian McDermott Dec 5, 2025 ▶ 35:15 Zendesk's Adrian McDermott: AI's Customer Service Potential, Adoption Cycle, Scale vs. Orchestration
Dec 5, 2025 negative
Insight
McDermott: Autonomous continual learning creates an unverifiable black box for enterprise support
“When we're changing the machine and we're writing new articles and we're building new macros, it's sort of easy for the humans who own that machine to understand what's going on. Like the head of support technology can be like, yep, yep, that is the correct wo…”
Adrian McDermott Dec 5, 2025 ▶ 37:30 Zendesk's Adrian McDermott: AI's Customer Service Potential, Adoption Cycle, Scale vs. Orchestration
Feb 25, 2026 negative
Assertion Not checkable as stated
Pollan: Computer scientists agree LLMs will not lead to machine consciousness
“I've asked other people who are pretty knowledgeable computer scientists. Nobody seemed to think large language models are the way to go toward consciousness.”
Michael Pollan Feb 25, 2026 ▶ 29:28 Can AI Become Conscious? — With Michael Pollan
Mar 16, 2026 positive
Assertion Supported
Ranjan Roy: Meta rebuilt its ad infrastructure using LLMs and agentic AI
“They rebuilt their entire advertising infrastructure to incorporate large language models and agentic processes and the newer vintage of AI rather than the traditional machine learning, and that's what's made it, I mean, it basically just, not saved, but like,…”
Ranjan Roy Mar 16, 2026 ▶ 14:10 AI Backlash Intensifies, Nvidia GTC Preview, Meta’s Embarrassing Delay
Apr 22, 2026 bearish
Opinion
Véliz: Large language models have no regard for truth
“And that's essentially what a large language model is. It has no regard for the truth. It wants to please you. If what pleases you happens to be true, great. But if it's not true, then it doesn't care one way or another.”
Carissa Véliz Apr 22, 2026 ▶ 35:53 Are We Too Obsessed With AI Predictions? — With Carissa Véliz
May 20, 2026 bullish
Insight
Cherny: Next-Token Prediction Forces AI Models to Plan Ahead
“It's crazy, like, you teach this thing to predict the next word, and somehow, if the next word is hard enough, it has to learn to really plan ahead, and it has to learn how to do all of this.”
Boris Cherny May 20, 2026 ▶ 54:20 Claude Code Head Boris Cherny: Insane Growth, Tokenmaxxing, AI Agents' Next Frontier
May 26, 2026
Insight
Enterprise AI requires wrapping probabilistic LLMs in strict, deterministic security harnesses
“There is an LLM that's reasoning and planning at the core. But there is a harness around it that provides a lot of determinism. And so that determinism is adding in governance, security, trust, integrations, permissions, and that's really what's going to make …”
Joe Davis May 26, 2026 ▶ 19:57 The Right Way To Build AI Agents — With NVIDIA's Adel El Hallak and ServiceNow's Joe Davis
Jun 16, 2026 bullish
Insight
Ghai: Automating document workflows speeds up hospital emergency intake and referrals
“If we can automate all of these document centric tasks at machine speed, patients get onboarded and assigned a bed in the emergency department much faster. Nurses are freed up to deliver care with more of their time. Referrals can happen at machine speed so pa…”
Jitesh Gai Jun 16, 2026 ▶ 19:58 Is Unstructured Data The Key To Successful AI Deployments?
Jul 7, 2026 negative
Assertion Supported
Kutylowski: General LLMs are predominantly trained on English and few other languages
“Large language models, those that are meant for generalized usage, they're usually trained mainly on English, and maybe a couple of languages, and if you really want to go broad, if you really want to go multilingual, you need models, you need the data and the…”
Jarek Kutylowski Jul 7, 2026 ▶ 5:30 Why Specialized AI Models Are Challenging the Frontier Labs — With DeepL CEO Jarek Kutylowski
Jul 15, 2026 negative
Opinion
Schmidhuber: Large language models alone cannot lead to AGI
“Yeah, so when the phrase, when the question is phrased like this, does an LLM, a large language model by itself, lead to AGI, the answer is a clear no.”
Jürgen Schmidhuber Jul 15, 2026 ▶ 8:40 AI Pioneer Jürgen Schmidhuber: AI Already Feels Pain, Loves, and Is Self-Aware
Aug 19, 2026 positive
Opinion
Bostrom: Current LLMs Arguably Have Higher Ethical Standards Than Most Humans
“Current LLMs, like you're using them as an ordinary person, for the most part, they are helpful, and they try to solve your task that you assign them, or give an answer that is, sometimes they hallucinate, or maybe deceive a little bit, but broadly speaking, t…”
Nick Bostrom Aug 19, 2026 ▶ 18:37 Nick Bostrom: Worries About AI Existential Risk Just Became More Concrete
Aug 19, 2026 positive
Assertion Supported
Bostrom: Anthropic research found global workspace structures in large LLMs
“And so there was a recent paper by Anthropic looking at the existence of a kind of global workspace inside these large language models. This is the idea of there being a kind of almost like a stage inside a mind where some small subset of all the information t…”
Nick Bostrom Aug 19, 2026 ▶ 47:44 Nick Bostrom: Worries About AI Existential Risk Just Became More Concrete
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.