You're making good points. The selection bias one especially. 
The corpus is AI safety-focused podcasts (AXRP, 80,000 Hours, 
The Inside View, etc.), so it skews toward people who already 
take alignment seriously. LeCun, Ng, and others who think the 
framing is wrong don't typically appear on those shows, and 
when they do they're pushing back on the premise rather than 
giving a number.

The "10 to 99%" framing was honest about what's in this corpus 
but you're right that it implies a complete range. The "MacAskill 
among the lowest in serious circles" line was sloppy. That 
phrasing came from his own framing on 80k Hours but it carries 
an implication I didn't intend.

If anyone has pointers to AI safety podcast episodes with people 
in the under-5% range or rejecting the framing entirely, I'd 
appreciate it. That's a real gap in coverage I should address.

Thanks for asking, honestly not sure why the early downvotes. 
Guess the framing read too promotional first time around. Karma's 
recovered since. Worth mentioning the mobile experience was broken 
back then; that's fixed now if it matters.

Thanks! The disagreements are what make this field so interesting to map , the gap between Yudkowsky and MacAskill isn't just about probability estimates, it's about fundamentally different models of how AI development unfolds.