Anduril is great, but the problem is that China has 100 Andurils.
POLICY
-
Clarifying regulations on covered AI models and commercial use
By
–
ok that's a fair question. I'll delete my tweet since it's less clear than I thought. e.g. this doesn't make sense afaict: "Before using a covered model or covered model derivative, or making a covered model or covered model derivative available for commercial or public use, the
-
Critical Harm Definition for Covered AI Models and Derivatives
By
–
"(g) (1) “Critical harm” means any of the following harms caused or enabled by a covered model or covered model derivative:" They're treated exactly the same way.
-
White House discovers intelligent life signs amid escape concerns
By
–
White House emergency: signs of intelligent life found. Will it escape? https://t.co/jVioAlL5SR
— Pedro Domingos (@pmddomingos) 10 août 2024White House emergency: signs of intelligent life found. Will it escape?
-
AI Model Threshold Regulations Expand Coverage to Fine-tuning
By
–
You misunderstand. The covered model threshold includes flops used in the base model, not just amount used for fine-tuning. So in a few years, nearly all significant models and all fine tunes of them will be covered.
-
Honesty essential for proper AI debate and discussion
By
–
Lying about it isn't reasonable. How can it be debated properly without telling the truth?
-
Open Source Developers Face Legal Risks From State AI Bill
By
–
True, but those behind the bill are claiming it covers anyone doing business in the state. I'm not sure that's true, or legal, but I wouldn't want to be the open source dev that has to deal with that in court.
-
SB1047 Blocks Open Source AI Fine-Tuning Development
By
–
I used to be in the "benefit of the doubt" camp, but can no longer — so lemme be clear:
this is quite simply a lie. Those pushing SB1047 *know* that nearly all open source AI devs rely on fine-tuned models, which *are* blocked by this bill. -
Regulatory Thresholds Could Eliminate Open Source AI Models
By
–
Once models have reached the FLOPs threshold in a few years, the mathematically impossible requirements this bill puts on devs means that no open source models could meet them, so we'll all be entirely reliant on big tech's commercial closed offerings.
-
Frontier AI Regulation Through Policy Red-Teaming Approach
By
–
My work on "Frontier AI Regulation" last year was based on this policy red-teaming approach 🙂