(Unlike nuclear power)..”there exists no scientific basis or evidence for how or when AGI will emerge (if ever), leaving us with only a hypothetical risk that has capitulated many of our regulatory efforts today. What we are seeing from current AI leaders on their comparison to
REGULATION
-
Stablecoin Regulation and Startup Investment Currency Risk
By
–
No stablecoins except those approved by the relevant regulator. (Currently empty set I believe, off top of my head.) If for some reason one wants to involve a slow database in one’s startup investing, one needs to take currency risk and (at least) two legs of fees.
-
Japanese crypto regulations limit USDC cash-out for startup investing
By
–
You may be surprised to learn that under the current regulations no Japanese crypto exchange will allow you to cash out USDC, which feels relevant to the claim of whether it is actually useful for investing in a Japanese startup.
-
Banks struggle with fraud detection and velocity limits
By
–
You would hope that banks would be able to heuristically detect this but a combination of organizational, IT, cultural, and regulatory issues makes that difficult, and so they adopt individualized velocity limits to minimize damage and attempt to educate users about this.
-
AI Labs Must Complete Nuclear Safety Analogy Seriously
By
–
"If AI labs are to consistently invoke exaggerated fears through comparisons to nuclear hazards, then they must be willing to take the analogy to completion. Ironically, if they were to explore the readily available safety mechanisms for nuclear components,…
-
Regulatory Policy Limits Startup Access Despite KYC Compliance
By
–
This is the motivation behind the policy but the policy does not allow them to e.g. read a share purchase agreement or make use of the fact that the startup is a real company and KYCed by the receiving bank so there is no risk to this sophisticated customer.
-
Generative Models and Dangerous Compound Synthesis Safety Concerns
By
–
I don't get it, personally. They "chose to drive the generative model towards compounds such as the nerve agent VX" — and the model did that pretty well. E.g it recreated VX. VX is bad – but there's already loads of public info on how to make it.
-

AI’s Role in Society: Generative AI Panel Discussion
By
–
Our last panel of the day focuses on envisioning AI’s place in society. Join @RishiBommasani
, @erikbryn
, @microsoft
’s Jaron Lanier, @joon_s_pk
, & @MIT
’s Ashia Wilson for a wide-ranging discussion at the HAI Fall Conference on new horizons in generative AI: https://
stanford.io/46NtzuT -
AI Safety Summit: Balancing Risks and Beneficial Opportunities
By
–
Ahead of the AI Safety Summit, I spoke with the @guardian about why we need to take the risks of AI seriously while not losing sight of the incredible beneficial opportunities
-
Senate AI Insight Forum participants and policy context revealed
By
–
Also, @techpolicypress published a list of today's participants and context about the first "AI Insight Forum" that took place last month: https://
techpolicy.press/us-senate-ai-i
nsight-forum-tracker//
…