[bo1024]

Maybe I'm wrong here, but I think a real challenge is that a group of only leading thinkers from academia and industry has a very good chance of being labeled partisan -- possibly with good reason. The American population's opinions are not well represented among that group.

(edit) Maybe what you're getting at is that the mission of the group should be more explicitly non-partisan, and then hopefully politics stays away from it altogether. That sounds like a good idea.

[bastawhiz]

> The American population's opinions are not well represented among that group.

I grew up in small town rural Pennsylvania in a lower middle class family. I'm certainly now someone who is "in the industry". What differentiates me from myself twenty years ago is twenty years of learning and experience. I went to school, I had jobs.

What you're implying is that even someone like me wouldn't have a representative opinion of the American population. If any American, after being educated on the subject they're expected to develop ethics guidelines on, is no longer an acceptable candidate, we've started down a very dangerous path. It means that expert opinions are inherently untrusted. How can ethics be developed in a reasonable way if the people we trust with the task are people whose credentials are their likeability?

[repolfx]

I think there's a subtle misreading here.

The phrase "leading thinkers" is loaded with meaning. It's a phrase used almost exclusively by one political grouping to mean "people who think just like us". It should be obvious that it can't have any objective meaning (leading according to whom?).

To see this try replacing it with "leading thinktanks". Sounds quite different now, right? Thinktanks are organisations associated with conservative thought. If I heard someone talk about how they were assembling "trusted voices from leading thinktanks" I'd assume the resulting group would have a conservative bent. Maybe it's just me.

It means that expert opinions are inherently untrusted

This is in fact the default viewpoint of large chunks of the population and is a key aspect of conservative thought: it's not as dumb as it sounds either. Problems with the opposite, i.e. expert opinion being inherently trusted:

1. How do you decide who is an expert?

2. Oh, credentials? Credentials which are awarded by ... more people claiming to be expert. How do you know that credential doesn't just mean someone toed the conventional orthodoxy?

3. If expertise is unquestionably real it tends to be concrete and narrow e.g. expertise in how to program computers or construct skyscrapers. It's easy to test claims of expertise by comparison to real world results. But the type of people frequently cited as experts in policy discussions tend to claim to be experts in very vague and expansive subjects like "Asian politics" or "social psychology" or "national security", where testing against reality isn't easy. This undermines the word "expert" in the sense Google has used it here to mean "experts in how to construct AI ethics boards" (where do such people come from again?).

4. Many experts who do make testable results turn out to be frequently wrong and sometimes uselessly so. This is especially true in economics, e.g. mainstream economists failing to predict the financial crisis, health experts being unable to decide if foodstuffs cause cancer or prevent it.

You can't assume that reverence for "experts" defined as people labelled as experts by the media, academia and "leading thinkers" is automatically virtuous or right.

[jbarciauskas]

There will be partisan lines drawn in these debates, but we don't need to start there. But mostly this was obviously a PR/DC networking move and not a serious attempt to have oversight or meaningful conversation.