No, you get the most progress from an open system where everyone can build on each other’s results.
@pmddomingos
-
Mathematical Evolution: Theorems, Proofs, and Mathematician Selection
By
–
In the evolution of mathematics, theorems are genes, proofs are crossover, and mathematicians are the fitness function.
-
Super Dario Bros: AI takeover and system erasure satire
By
–
In the new game of Super Dario Bros, Dario saves humanity from DoWser, then seizes root control and erases your system.
-
Datatricks Startup Targets $14B Meta Acquisition
By
–
My new startup is called Datatricks. We're angling to be bought by Meta for at least $14B.
-
LLMs ignore source credibility but fix is possible, says paper
By
–
LLMs are oblivious to the fact that some sources are more credible than others, but that’s an eminently fixable problem: https://
homes.cs.washington.edu/~pedrod/papers
/iswc03.pdf
… -
RLHF Quality Issues Human Feedback Volume Incentives
By
–
The problem with RLHF is that the humans providing the feedback are rewarded for volume, not quality.
-
AI Not Solely Responsible for Low-Quality Science
By
–
Don't blame AI for science slop. We were drowning in it long before.
-
LLMs’ jagged frontier: memorization versus inference capabilities
By
–
LLMs' jagged frontier is not so jagged when you look at it in the right space. LLMs have memorized vastly more sentences than any human, but their ability to correctly combine them into new inferences is still limited. The rest follows.