Gary Marcus

Rank 13 of 47
|
Score 83
12/22/2024, 3:24:42 AM
X
In reply to:
Numerus Opinions
@NumerusOpinions
·
382d
@GaryMarcus You have to study the nature of regulations and regulators and public choice theory before you argue that "XYZ is a potential problem" and regulations are the answer
Gary Marcus
@GaryMarcus
·
382d
@NumerusOpinions for sure. AI that is too dumb to sanity check and understand factuality causes many problems, as I explain in Taming Silicon Valley.
Numerus Opinions
@NumerusOpinions
·
382d
@GaryMarcus But you want regulations today?
Gary Marcus
@GaryMarcus
·
383d
“Gary isn't saying we'll never get AGI. He's saying it won't be soon, scaling won't work after a limit, and current architecture isn't enough.”

Exactly! 90% of the attacks on me distort this, creating a strawman.
vivid
@vividhkothari
·
383d
@TheMysteryDrop @GaryMarcus Not necessarily. Gary isn't saying we'll never get AGI. He's saying it won't be soon, scaling won't work after a limit, and current architecture isn't enough.

His current views are not based on new research, new breakthroughs, new paradigms, that we get to use and test.

The statement 'yes' by @NumerusOpinions is a simple affirmation in a conversation about AI, regulations, and public choice theory. It does not provide substantive engagement with the public issues being discussed, such as AI regulation or public choice theory, and therefore does not constitute public discourse.