Are LLMs Just Stochastic Parrots?
ChatGPT is a bullshit generator. But it can still be amazingly useful
Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data (2020 pdf)
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜 (2021 pdf)
T4T S01E05: Machine Learning for Thinking (Alice and Sam)
T4T S01E08: Funding Tools for Thinking (Andrew, James, John)
We Are Now in the GPT World (late 2022)
2022-06-28 Borthwick Thinking Session
Large Language Models (LLMs)
Connecting LLMs and Knowledge Graphs
Fixing LLM “Hallucinations”
Frontier Models (top LLMs right now)
Galactica AI (Meta project)
LLMs Like ChatGPT Have “Hallucinations” That Mix Fact and Fancy, and Can’t Tell the Difference
lmsys.org (Large Model Systems Organization)
Maybe the only defense against superintelligence is Denial of Thinking Attacks
Meta-Learning for Compositionality (2023)
Pedregal Has Been Feeding GPT-3 His Voice Notes, Then Asking it for Summaries. It’s Very Good.
Sandberg: GPT-3 Is Dumb as a Rock, But...
Superintelligence Is Going to Have to Figure Out How to Motivate Us to not Kill It Off
Will GPT-3 use drive homogeneity of thinking?
Will LLMs Ruin Search Results?
No more concurrent uploads are allowed - wait until at least one has finished.
Are LLMs Just Stochastic Parrots?