Periodic public service announcement, which unfortunately all too often seems necessary in this joint: Always aspire to the top, not the bottom, of @paulg
’s beautiful pyramid of argumentation:
@garymarcus
-
Aspire to quality argumentation in AI discourse
By
–
-
Architecture and Investment: Understanding AI Subtleties
By
–
its a matter of architecture and investment, not semantics. we will get there sooner if people better understood these subtleties.
-
Claude Code Demonstrates Neurosymbolic AI Beyond Pure LLMs
By
–
Claude Code is suprisingly good. BUt also pretty clear neurosymbolic, not a pure LLM (which is what i have advocated for decades). my beef is with pure LLMs, not neurosymbolic approaches.
-
Media coverage of changing AI estimates and forecasts
By
–
not a fair criticism of Daniel (who is just regularly updating things to give his best estimate in light of changing data), but fair criticism of how media often reports this stuff
-
Toasters and Superintelligence: Debunking AI Hype
By
–
and toasters consistently produce value to millions of people, so they must be superintelligence!
-
Free AI Models Misleading Public on True Capabilities
By
–
this is disingenuous. for many people the models are the free models. massive public harm has been caused by leading people to believe that the free models are something they are not. i hope we can agree on this.
-
AGI Definition Debate: Generality vs Narrow Problem Solving
By
–
this is soooo wrong. the whole point of AGI is to be *general*, across all problems. that’s literally what the G stands for, general. Olympiad problems are narrow, and don’t per se have anything special to do with AGI. For some discussion of AGI definitions, see @hendrycks et
-
Media Hype Over AI Predictions and Revised Estimates
By
–
to be clear, i think it is great that he updates his estimates regularly based on new data. all scientists should do that. my irritation is with the media that hypes up the early predictions and delta forward and pays less attention to the later ones and when the delta goes
-
Media Responsibility in AI Predictions and Coverage
By
–
not penalizing Daniel; penalizing the media (or in this case) Grok. Daniel is a good fellow who does his best to make careful predictions. (we had a great debate on this stuff @MLStreetTalk
) -
Maybe: The Next Decade in AI – 2020 Predictions
By
–
Maybe: The Next Decade in AI, free on arxiv in 2020, anticipated a lot of where we are now.