Correct. Hinton is brilliant and we owe him much gratitude.
@jeremyphoward
-
Phi 1.5: TinyStories for Code Release
By
–
phi1/1.5 is tinystories for code. Includes some of the same authors.
-
First Principles Thinking for AI Intervention Analysis
By
–
Don't look for things that just sound similar and assume you can borrow what you know from there. Think deeply, from first principles, and considering beyond 1st order effects, about what interventions will do.
-
Frontier Model Licensing Proposals Car Analogy Critique
By
–
Your attempt to align frontier model licensing proposals with car licensing is confusing two things that happen to use the same words but are *totally* different. This is a lazy analogy.
-
Misinterpreting AI Risk Discourse: Assumption versus Reality
By
–
You say it's important "we *also* talk about the very real risks that many people… see and believe we need to mitigate"; and yet, I explicitly say lets *assume* those risks are real in the first minute of my talk. You're responding to what you assume you hear, not what's real
-
70B LLM Inference Optimization in Kaggle Competitions
By
–
Yes exactly that – although it's particularly used by Kagglers in competitions, since many of them *require* using Kaggle's infra. In a recent competition one person got 70B inference running in 14GB RAM! :O
-
Regulatory Capture Worsens AI Concerns Rather Than Solves
By
–
Of course concerns are warranted. But regulatory capture is not a solution to those concerns. It makes them worse.
-
Knowledge Restrictions in AI: Ethics Versus Governance
By
–
That's highly controversial, and attempts to minimize the spread of some kinds of knowledge have largely been consigned to history (and a handful of autocratic states).
-
Project Manager Nuclear Risks Expertise and Accountability Question
By
–
The project manager for the Nuclear Information Project is outside the field of commenting on nuclear risks? You do realise that the quoted tweet is explicitly about that topic, right?
-
Training Strong Models with Limited CPU and RAM Resources
By
–
This is an answer to your question "Have any of you run into compute cloud providers that had particularly weak CPUs?" It's an interesting platform for seeing how people handle extremely low RAM/CPU availability and still train strong models. It's impressive what people manage!