Hi, I’m Max :)
working in AI governance (strategy, expert surveys, research infrastructure, EU tech policy fellow)
background in cognitive science & biology (did research on metacognition and confidence judgements)
most worried about AI going badly for technical & coordination reasons
vegan for the animals
doing my own forecasts: https://www.metaculus.com/accounts/profile/110500/
Is it possible to switch to remote work (in my case in Europe) after working from office in SF e.g. for one year?
Thank you for this!
I wonder if EAAs might think that the main effect of their animal advocacy is not necessarily the immediate meat consumption change, but more the longer term moral circle expansion. For example Jacy Reese from the Sentience Institute recently argued in this direction. If so, it might be more interesting to test for the level of speciesism or level of empathy towards animals before and after those interventions. I’d guess that decreasing speciesism is easier than causing a change in consumption habits, so for moral circle expansion purposes those results could be encouraging, correct?
On your grants page (https://www.effectivealtruism.org/grants/) it still says you consider funding education, has this changed again?
“We welcome applications in the following areas: [...]
Studies (e.g., Masters or PhDs)”
The speciesism scale that was recently published by an EA psychologist might be useful for this purpose.
Ha, cool, you were also involved in the Sentience Institute study about attitudes towards US farm animals that probably also contains useful methodology in this direction.
Thanks for the summary! I don’t know if that came up during your discussion, but I would have found concrete examples useful for judging the arguments.
“ii) This causes EA to prioritise building relationships with high-status people, such as offering them major speaking slots at EA conferences, even when they aren’t particularly rigorous thinker”
I’d hope that bad arguments from high-status people will be pointed out and the discussion moves forward (e.g. Steven Pinker strawmanning worries about x-risks).
“iii) It also causes EA to want to dissociate from low-status people who produce ideas worth paying attention to.”
For example I find it unlikely that an anonymous writer with good ideas and comments won’t be read and discussed on the forum. Maybe it’s different on conferences and behind the scences at EA orgs, though?
“iv) By acquiring resources and status EA had drawn the attention of people who were interested in these resources, instead of the mission of EA. These people would damage the epistemic norms by attempting to shift the outcomes of truth-finding processes towards outcomes that would benefit them.”
EAs seem to mostly interact with research groups (part of the institution with the best track record in truth-finding) and non-profits. I’m not worried that research groups pose a significant threat to EAs epistemic standards, rather I expect researchers to 1) enrich them and 2) be a good match for altruistic/ethical motivations and being rigorous about this. Examples that come to mind are OpenPhil causing/convincing bioriks researchers to shift their research in the direction of existential threats.
Does someone know of examples or mechanism how non-profits might manipulate or have manipulated discussions? Maybe they find very consequential & self-serving arguments that are very difficult to evaluate? I believe some people think about AI Safety in this way, but my impression is that this issue has enjoyed a lot of scrutiny.
I believe that the feeling of loneliness probably is one big contributor to mental health issues and I like your idea of tackling it pragmatically/for-profit.
My gut feeling is that this won‘t be used and people are happy enough with the Craigslist solution. Anectdotally, my roommates (intelligent people) thought it was a joke when I proposed that I would design a questionnaire for people applying to live with us. Another platform, OkCupid, tries to offer meaningful matching scores for romantic partners and it seems to be rather fringe, at least in Germany.
I appreciate you seeking feedback here. A book targeted at the general public seems very well placed to shape the discussion in many unintended ways.
Evidently, the “AI kills” part is designed to be clickbait. We do hope to reach a wide audience, which we regard as desirable to accelerate the spread of AI ethics in all sorts of corporations. Evidently, we expect the title to be misused and abused by many people. But we are confident that a 3-minute discussion with a calm person is sufficient to convince them of the relevancy of the title (see Chapter 3).
I wonder why you think the sensationalist title is worth it. Less sensationalist title will probably cause
less unproductive debate
lower risk for unintendedly causing the discussion around AI to go down a confused and politicized path
less people will read it
but the people that will read it are probably more informed and interested to have a sober discussion
Can you explain why you think that AI ethics being discussed in all sorts of corporations is very useful? My impression is that AI Safety mostly needs more academic talent directed to research, and not being another hot topic in corporations that don‘t even do AI research.
Chapter 7 discusses the risk of AI race, and the constraints that this AI race implies. We argue that, because of this inevitable race, we cannot demand too constraining constraints on AIs.
I‘ve heard arguments that one wants to be careful when talking about AI races, as quotes like yours arguing about an inevitable race might fuel race dynamics further. Race dynamics are worrying and we should discuss the risk, of course. But maybe a book for the general audience, which I expect will jump on issues like the associated political conflicts as they are much easier to understand than technical challenges, might not be the best place for this discussion.
This received some discussion on LessWrong last year.
https://www.lesswrong.com/posts/2pnvgwGkMZkQzCvwi/a-future-for-neuroscience
Thanks for the update, I appreciate the transparency on the project’s shortcomings.
“Upon my initial review, it had a mixed track record. Some grants seemed quite exciting, some seemed promising, others lacked the information I needed to make an impact judgment, and others raised some concerns.”
I’d be interested in what (kind of) grants you think seem great and not so great.
I feel frustrated by the lack of feedback. The EA Hotel seems to be one of the most discussed projects born out of the EA community in the last months and it prominently struggles for funding. I think I’ve read most of the related discussions on the forum and haven‘t seen a case made why the project isn‘t as promising as it might sound after reading the EA Hotel funding pledges. I understand that your time is limited, but for the sake of cooperation within EA I’m saddened that there seems to have been no communication between you guys at all. :l
Thanks a lot for this, I found the post very informative and look forward to the next one.
One note: I would have found a more extensive list of countries in the geography section useful.
Thank you for writing this, I find the idea very interesting.
Your argument against worrying about negative consequences didn’t pick me up.
What if increasing these supposedly positive traits results in negative consequences?
It’s important to avoid status quo bias. To quote Bostrom and Ord:
“Reversal Test: When a proposal to change a certain parameter is thought to have bad overall consequences, consider a change to the same parameter in the opposite direction. If this is also thought to have bad overall consequences, then the onus is on those who reach these conclusions to explain why our position cannot be improved through changes to this parameter. If they are unable to do so, then we have reason to suspect that they suffer from status quo bias.”
I take issue with the Reversal test because it seems very difficult to have a realistic picture of a world where a parameter like IQ is increased or decreased significantly. Therefore, the “status quo bias” might be a reasonable attitude if we have very little understanding of the dynamics of our system and there is little chance of reversing the intervention (e.g. you mention a slippery slope dynamic where once genetic enhancement is used everyone will more or less have to use it).
Imagine that I currently feel more or less content with the broad trends of humanity and would like to be very careful before supporting profound interventions like widespread availability of genetic enhancement.
Thank you for the update, and for your work.
I found this post very useful to think about my own career, thanks for writing it up. My prospects also don’t fall neatly into the top recommended paths, so I’d be interested in more discussion how to train my “good judgement”.
Summarizing your ingredients of good judgment:
Spotting the important questions (e.g. what do I need to learn to improve my decision the most?)
Having good research intuitions (good quick guesses, think critically about evidence)
Having good sense about how the world works and what plans are likely to work.
Knowing when they’re out of their depth, knowing who to ask for help, knowing who to trust.
What do you think about participating in a forecasting platform, e.g. Good Judgement Open or Metaculus? It seems to cover all ingredients, and even be a good signal for others to evaluate your judgement quality. When I participated in GJO for a couple of months, I was demotivated by the lack feedback for the reasoning in my forecasts. I only could look at the reasoning of other forecasters and at my Brier score, of course.
P.S: Your thinking appears to be very clear and you appear rather competent, so I wonder if your bar of “good enough judgement” to reasonably pursue non-standard paths is too high. I also wonder if people whose judgement you trust would agree with your diagnosis that you wouldn’t have good enough judgement for a non-standard path.
Thanks for this. I like the ribbon plots!
Did you by chance look at cause prio differences between countries and saw anything interesting? I dimly remember there used to be a trend along the lines of a bit more animal welfare in continental Europe, global poverty in UK and x-risk in the US.
Congrats for your first post! I think it’s well written; I like the structure of „problem-proposed solution-possible issues“, you write short and clearly, and you stated what kind of input you want from the community.
It was useful for me that you provided the example of meat eating as a coordination problem. I would have found more examples even more useful for thinking about the potential applications where a coordination platform is among the most promising approaches (btw, I think for meat eating it is not among the most promising).
I like your idea, but I’m also worried about your 2nd issue: nobody will use it. It seems to me like people are just not motivated enough by being a part of improving the world. Meat eating seems like a case in point: there is already a veggie community you can be part of (at least in Germany in every bigger city) and the marginal impact you have doesn’t even depend that much on coordination. Still it’s a tiny movement.
I think it’s reasonable that you are trying to think about the landscape and bottlenecks of behavior change and coordination before moving to action. There is probably much more to learn. For example, I’ve read this short report about change platforms in the context of changing organizations, that seems to have some success stories and learned lessons that are also relevant for you. This might be a much more tractable pathway if there are smaller scale important coordination problems. https://www.mckinsey.com/business-functions/organization/our-insights/build-a-change-platform-not-a-change-program
I just looked it up, you’re right. Here the full quote:
F. Bailey Norwood and Jayson L. Lusk, Compassion, by the Pound: The Economics of Farm Animal Welfare (New York: Oxford University, 2011), 223.
Even though I was pretty actively reading the forum in the last months I’ve missed one of the posts and all of the really great comments, so thanks a lot!
I’m wondering if there is some reasonable way to search for highly upvoted comments that were made after reading a post. The forum seems to keep track which comments were made after a user last opened the forum post, so maybe one could sort those comments by their upvotes? Or maybe by relative upvotes so it is not dominated by the most popular forum posts.
Nick Bostrom, Existential Risk Prevention as Global Priority