successful governments tend to be * more high-entropy, making it hard to remove or capture the checks-and-balances easily
* or tend to be for small populations and transparently run
@soumithchintala
-
Government Resilience: High-Entropy Systems and Transparent Small-Scale Administration
By
–
-
Corporate Capture and AI Regulation: A Paradox?
By
–
if there's a world where any or most regulation is captured by "corporate-driven government corruption" as you quote — then is proposing further regulation good or bad?
-
Smart Regulation Signals from Non-Incumbents and Consumers
By
–
on the other hand, if a non-incumbent or a consumer is asking for regulation, and they are smart, connected and knowledgable, then hear closely what they have to say — it probably has a lot of signal because they are incentivized to break up the consolidation of power.
-
Decentralization vs Centralization of Power in AI Systems
By
–
There are very few instances where centralization of power resulted in good, and countless instances where decentralization of power resulted in good. It's almost impossible to interpret a positive intent for someone's call for centralization of power if that person is smart,
-
AI Regulation Position: Analyzing the Regulatory Video Debate
By
–
If you think Regulation is good, and AI Regulation is essential, what is your position on this video? https://t.co/eItoXepugS
— Soumith Chintala (@soumithchintala) 16 septembre 2023If you think Regulation is good, and AI Regulation is essential, what is your position on this video?
-
Can we probe LLM reasoning abilities like biological minds
By
–
the question was intentionally ill-posed but its hard to agree with the notion that because it's algo/math, you cant probe into its reasoning or communication abilities — aren't we just blobs of matter.
i didn't give the LLM agency, i merely posed the question in an -
Are LLMs Unintelligent or Non-Native Language Users?
By
–
if LLMs are weirdly unintelligent, is it because they are stupid, or is it because they’re not a “native speaker”?
-
PhD Deep Learning: Unnecessary in Mainstream AI Era
By
–
i think you're saying — reject the notion of a PhD or a research Masters in hot deep-learning topics because AI is now so mainstream and so well-capitalized and so accessible that you can do great work on discord or in an industrial lab and publish and you win at life.
That -
Academic Incentive Systems and Performance Metrics in AI Research
By
–
very reductive take. a lot of people play the game as an essential side hustle or as a meaningful checkpoint of progress (academia is naturally unstable, students are only around for a little bit).
its okay, it’s not the worst.
its like the crap people do for perf reviews, like -
Doug Lenat’s Cyc Legacy: Symbolic AI Pioneer’s Ontological Work
By
–
AI lost a monumental scientist in Doug Lenat.
His work on Cyc deeply inspires me till today — not so much because he worked on symbolic AI but because of how far he'd taken it, through a multi-decade dedication.
I hope the propreitary Cyc work (ontologies, etc.) get open-sourced