large language models
AI systems trained on vast text data that can generate and analyze language, raised as a potential tool for historical research.
As defined in cwt_alan-taylor
Other definitions
AI systems trained on large text corpora that generate and process natural language; here credited with expanding research prep capacity.
Conversations with Tyler 2025 RetrospectiveAI systems (like ChatGPT) trained on massive text corpora to predict statistically likely next words; excel at reproducing probable average discourse
Nassim Nicholas Taleb: Connectivity, Global Fragility and the Added Danger of AI
Where it comes up · 43
- This is not the AI we were promised | The Royal Society17 claims
- Don’t Believe AI Hype, This is Where it’s Actually Headed | Oxford’s Michael Wooldridge | AI History8 claims
- Richard Sutton – Father of RL thinks LLMs are a dead end5 claims
- 174 | Tai-Danae Bradley on Algebra, Topology, Language, and Entropy5 claims
- 339 | Ned Block on Whether Consciousness Requires Biology5 claims
- Richard Sutton – Father of RL thinks LLMs are a dead end4 claims
- Geoff Hinton - Will Digital Intelligence Replace Biological Intelligence? | Vector's Remarkable 20244 claims
- Lex Fridman Podcast #428 - Sean Carroll4 claims
and 35 more
Most important claims · 10 of 110Search all claims for “large language models” →
- Machine translation historically was only useful for economically important languages because of the corpora used, but new large-language-model methods have dramatically improved translation for smaller languages like Zulu, which will let small-language communities produce shareable articles and accelerate the largely invisible growth of Wikipedia in developing-world languages.Jimmy Wales
- Large language models produce text that is not only coherent—with logic and structure across whole paragraphs and texts—but convincing, even when it gets facts wrong, revealing that the human brain loves language so much that we don't need the facts to be correct, only the story to be beautiful, which has been the secret of politics and religion for thousands of years and now comes with AI.Yuval Noah Harari
- Philosophical questions about AI (does it think? is it conscious? does it understand?) have shifted from pure philosophy into experimental science because we now have systems to test; questions that were purely metaphysical are becoming empirical investigations.Don’t Believe AI Hype, This is Where it’s Actually Headed | Oxford’s Michael Wooldridge | AI History
- People interact with large language models as if they were rational minds because humans are evolved to seek out rational minds and interact with them, and LLMs are designed to communicate in very natural ways that trigger this evolved response.This is not the AI we were promised | The Royal Society
- We can now understand a gene's function through multiple facets: literature via large language models, knowledge graphs showing disease interactions and chemicals, protein structure via protein language models, expression patterns across datasets, and protein-protein interaction networks.MIT Scientist on Unifying Cognition and Biology | Manolis Kellis
- What we want, following Alan Turing, is a machine that can learn from experience, where experience is the things that actually happen in your life—you do things, you see what happens, and that's what you learn from, not learning from 'here's a situation and here's what a person did.'Richard Sutton – Father of RL thinks LLMs are a dead end
- Because large language models successfully reproduce language using only algebra and statistics, some mathematical structure must be learned inside the neural network, and the open research question is to identify and characterize that underlying mathematics.174 | Tai-Danae Bradley on Algebra, Topology, Language, and Entropy
- We cannot directly inspect another being's internal model of the world; just as we can only ask a person what their model is rather than open their brain, the only accessible test of intelligence or understanding is its outward phenomenology, which is what makes the Turing test framing compelling.335 | Andrew Jaffe on Models, Probability, and the Universe
- There is more continuity between biological life and technological life on Earth than between Earth's biological life and a different alien instantiation, because there is a direct lineage of information from the last universal common ancestor up through modern computers—for example, DNA can be sequenced, read into a computer, and become a computer virus, and large language models have the structure of human language precisely because human minds were the environment that selected them.388 - What Is Life
- The most promising criterion for AI consciousness is that an AI trained on data with all first-person point of view removed (e.g. only the Encyclopedia Britannica) that nonetheless spontaneously developed a first-person point of view would be far more convincing evidence of experience than current LLMs, which are trained on vast human-generated first-person material.339 | Ned Block on Whether Consciousness Requires Biology