I'm going to be interviewed by Tyler Cowen for "Conversations
With Tyler" in a few weeks. Tyler has asked
the public for suggestions for questions he should ask me.
Initially to amuse myself, but then with increasing seriousness, I
spent a couple of hours brainstorming a list of such questions.
The list is below, for posterity. Also: at the end a list of
observations about podcasts, and space for you to suggest
questions.
The Great Stagnation is often associated with bad effects on
our society – declining productivity growth, declining real wages,
declining impact of technology. What, if any, positive impacts
does the Great Stagnation have? [This is also a question I'd like
to ask Tyler; his answer might well be more interesting than
mine.]
You've recently been writing about AI and AGI and xrisk. These
are worthwhile subjects, but they're also extremely popular, and
it seems as though you're following the herd. Much of your earlier
work was in areas which were unfashionable at the time, like
quantum computing in the early 1990s, and which only became
popular later. Why have you joined the large group of people
opining on AI xrisk? Can't you find better ways to
contribute?
The physicist Richard Feynman once said that he didn't
understand quantum mechanics, and neither does anyone else. What
did he mean by that? Has the human race made much progress on the
problem since then? Have we solved it? If not, what are some
promising strategies to make a dent in it?
Should the founder of SciHub, Alexandra Elbakyan, receive the
Nobel Peace Prize?
You're a fan of the work of the physicist Alexei Kitaev, and
have said you think it's of epochal significance. Can you explain
what his work is, and why you think it's important?
There's a pretty thin line between dilettante and polymath.
Specialists often do the best work, simply because they put more
time into it, and can become far more knowledgeable and accrue
more resources; they therefore have many advantages in arriving at
the frontier and then expanding it. How can someone pursuing
multidisciplinary work ensure they're not just faffing around, and
actually contribute in a way that specialists cannot? In a related
vein: your early work was in the hard sciences, involving lots of
complex mathematics. You've gradually moved toward much softer
areas of work. Do you think such work is as valuable? What are the
pluses and minuses of that shift?
You were asked if you'd be interested in joining OpenAI when
it was starting up in 2015. Do you regret going in another
direction?
People like Steve Weinberg and Stephen Hawking have talked of
physics nearing a theory of everything. Does this mean physics is
nearing an end? Will attention necessarily shift to disciplines
like biology? Will they be more productive?
Your Twitter bio reads "Searching for the numinous". What does
that mean? Why did you choose that? How do you think people can
improve the way they self-describe? How is it related to "raising
others' aspirations"?
You began to work on open science advocacy around 2006, and
did it fulltime from 2007 through 2012 or so. How much progress
has been made an open science since then? How much creative
ferment in the media scientists use to communicate their work?
What is the opportunity here, to restructure expert
attention?
The Alignment Problem is often described as the problem of
aligning computer systems – especially future ASI systems – with
human values. You've proposed a very different framing of the
problem – something like: "How to align a liberal society so it
very strongly prioritizes differential technology development,
while preserving liberal values?" Why make the switch? What's this
got to do with AI? What benefits does the new framing give? How do
you think humanity is going to solve it? What role do you think
surveillance is likely to have in a solution? Can that be done
without creating an authoritarian state?
Why hasn't more progress been made on molecular
nanotechnology?
You initially were skeptical of Emergent Ventures. But you've
changed your mind and become enthusiastic about it. What caused
the switch? What would you change about Emergent Ventures if you
were running it?
And, for context, a couple of things I believe about
podcasts:
Most podcasts are boring because the interviewer is asking
questions they already know the answer to, faking the role of an
ignorant audience member. They're much better off asking questions
they don't know the answer to, but
genuinely want to know the answers to. One such
genuine question, of shared interest, can often support many hours
(or even a lifetime) of back and forth; but if they have no such
questions, they have the wrong guest. This has many consequences.
One is that they sometimes must explain a fair bit of background
context for the audience. Second, it means that a well-prepared
interviewer should not be asking questions designed to get their
guest to restate things they've published – those ought to be
presumed common knowledge. Rather, they will be asking frontier
questions – clarifying and pushing beyond what's been published.
This ought to feel exploratory, a bit wild.
The most interesting questions for any person are often the
questions they don't know to ask themselves, things in their
creative blind spot. It follows that an interviewer ought to ask
me much better questions than those above. But perhaps the
questions are nonetheless useful as stimulus.
