I don't have anything in mind, but since some people like you assimilate technology like AI and synthetic biology with powerful weapons, it would only seem logical to push the analogy: we don't put powerful weapons in the hands of anyone without testing their mental stability.
SAFETY
-
AI Hallucinations: Challenges to Innovation and Goal Alignment
By
–
AI's potential for innovation is vast but not without challenges. "Hallucinations," or nonsensical data produced by unsupervised AI, can lead to unexpected results that stray from the set goal.
-
AI Design Philosophy: Enhancement Not Destruction
By
–
Well, that was kind of the idea behind the Mutually Assured Destruction doctrine, but nevermind. AI is designed to make humanity smarter, not wipe out entire cities.
I fail to see how this analogy has any relevance. -
Does AI Regulation Increase or Decrease Extinction Risk?
By
–
Geoffrey do you think advocating for strong regulation will make human extinction more likely, or less likely? On what basis do you make that assessment?
-
AI Safety Policies May Undermine Protection Against Superintelligence
By
–
AI Safety advocates are *not* really taking seriously the possibility of AI super intelligence. If you do really think that this is a danger, their main policy recommendations may make us much *less* safe. I explain why in this discussion:
-
AI Extinction Risk: Human Agency and Potential Salvation
By
–
My estimate is:
"considerably less than most other potential causes of human extinction"
Because we have agency in this. It's not like some sort of natural phenomenon that we can't stop. Conversely, AI could actually save humanity from extinction.
What is your estimate for that -
White House Defense Production Act Distorts AI Security Policy
By
–
Laws to ensure AI applications are safe, fair, and transparent are needed. But the White House's use of the Defense Production Act—typically reserved for war or national emergencies—distorts AI through the lens of security, for example with phrases like "companies developing any
-
Nuclear Weapons Analogy Relevance to AI Safety Concerns
By
–
Geoffrey, that's a really weak analogy. I'm surprised you think it's relevant. Can you explain on what basis you feel like your understanding of nuclear weapons is relevant to the use of AI, and why the differences between the two are not relevant?
-
Deep Learning Models vs Human Cognition: Recitation, Intuition, and Reasoning
By
–
DL models do chiefly 2, with some 1 in the mix. Some rare AI systems attempt to do 3. Humans do a combination of 2 and 3, for the most part (recitation is a thing but is les common). By "volume", most cognition is intuition, but reasoning accounts for the most critical bits.
-
Liability Claims in AI Policy Summary: Verification Discrepancy
By
–
The word "liability" doesn't appear AFAICT on the web page, arxiv PDF, or policy supplement — but it is in your 5 point summary. Where are you getting that from? (Sorry if I'm missing something obvious!)