The real AI danger is all the loons it’s unleashing.
— Pedro Domingos (@pmddomingos) 1 juillet 2024
pic.twitter.com/6zAdKPCkyZ
The real AI danger is all the loons it’s unleashing.
@pmddomingos
-
The Real Danger of AI: Unleashing Unstable Individuals
By
–
-
Brute Force Limitations in AI Development and Jensen Huang’s Perspective
By
–
Brute force will not solve AI, but Jensen Huang wants us to try.
-
Generative AI Compared to Real Artificial Intelligence
By
–
Generative AI is to real AI like these gymnasts are to real ones.
-
Multimodal Models and Hallucination Problem Skepticism
By
–
And they told us multimodal models were the cure for hallucination.
-
Stock Market’s Multitrillion-Dollar Bet on Major AI Progress
By
–
The stock market is making a multitrillion-dollar gamble on major progress in AI. No pressure.
-
Ne pas confondre interpolation avec intelligence
By
–
Don't confuse interpolation with intelligence.
-
AI excels at EQ and creativity more than IQ and reliability
By
–
Larry has it exactly backward. EQ is much easier for AI than IQ, just like creativity is much easier than reliability. https://t.co/jWg2uSaaCM
— Pedro Domingos (@pmddomingos) 29 juin 2024Larry has it exactly backward. EQ is much easier for AI than IQ, just like creativity is much easier than reliability.
-
Biden Compared to Low-Order N-gram Model Humorously
By
–
If Biden was an n-gram model, his n would be pretty low.
-
Scale AI’s Ironic Business Model: Data Labeling Over Scalability
By
–
In keeping with the rule that AI companies' names are the opposite of reality, Scale AI does the least scalable part of AI (data labeling).