timtyler comments on Q&A with experts on risks from AI #4

[timtyler]

I think the answers to Q1 should resemble a log-normal distribution.

David Lucifer apparently agrees. I notice that not all interviewees seem to agree, though.