I've been researching Responsible AI/AI Safety and recently started trying to map the space. My research didn't turn up any non-stale "awesome-" lists on GitHub except for one, which was owned by someone who also owns hundreds of other awesome lists and who has included a lot of donation/support-me CTAs. There was some decent content, but overall it looked like a land-grab.
So I decided to create an awesome-responsible-ai list for humans to curate:
In the spirit of the OP's post, I propose a new post: does hiring Impact Analysts work? You might want to commission a study of impact analyst effectiveness in real-world situations that involve small and messy data with long durations to outcomes.
While an eager Impact Analyst might nerd out on all sorts of fun academic tools and theoretical proposals, in industry and at this stage of your growth, nobody should really care about measuring if your programs work and everyone should share about making them work as well as possible. Spoiler: they work. If you don't accept that, then you have to accept that your programs certainly aren't hurting anyone and shouldn't be prematurely derailed because someone is skeptical about whether or how well they work.
IMHO, define a short-term performance metric and put this budget toward a relentless and talented PM who is incentivized by the metric. They should be able to set up the basic data gathering (you don't need anything fancy), listen to customer feedback, and iterate on your product quickly.
Thank you for sharing this. I largely agree with your views on how EA should evolve, but my main question is: if it did, would it still be EA and not just some kind of privately funded and significantly smaller version of the external ecosystem that already contains researchers, multilateral aid organizations, and governments? Is an evolved EA just a microcosm of what already exists elsewhere?
So what options does EA have?
It can evolve its values, identity, and methods, in the ways you've recommended. But in this case I would argue it's just altruism and not much different than the organizations and ecosystem it originally claimed it was more effective than.
It can stay true to its original identity, but in doing so continues to be constrained.
Could it be that the EA brand (but not the community!) has become obsolete?
I find it encouraging that CG are investing in things like capacity building and relying on anecdotal evidence as insight. Ten years ago, my impression is this would have been almost EA-heretical in the face of "gold-standard" RCT canon. Reading this article, I'm more convinced that high-impact opportunities do exist to build and bridge AI Safety communities, and that it's a meaningful direction to contribute to. I can't help but think that there are unaddressed niches in uniting a relatively fractured ecosystem across industry, government, and research. Practitioners, policy-makers, and researchers are very important roles, but they are only part of the equation as anyone familiar with organizational behavior knows. Here's to the generalists who can also make many things happen :)
Congrats on the funding! And thanks for sharing so much detail on your vision and strategy. I've been very interested in contributing to this space and have been speaking with a variety of international university groups over the past few months about exactly these types of initiatives (at a more grass roots level). I'd be happy to help connect you with them!
Hi everyone, I'm Roland. I've worked in econometrics and experimentation for tech and consumer companies for many years and have been lurking on EA for the past couple. I'm a bit of a polymath, with a couple graduate degrees in computational physics and business administration. I'd like to engage more with people who are interested in working on effective and novel solutions to problems like extreme poverty, AI safety, and health. I am action-oriented and interested particularly in exploring low-friction, big-if-true POC solutions that can be trialed in these domains. I love meeting new people so feel free to say hi :)
awesome-responsible-AI
I've been researching Responsible AI/AI Safety and recently started trying to map the space. My research didn't turn up any non-stale "awesome-" lists on GitHub except for one, which was owned by someone who also owns hundreds of other awesome lists and who has included a lot of donation/support-me CTAs. There was some decent content, but overall it looked like a land-grab.
So I decided to create an awesome-responsible-ai list for humans to curate:
https://github.com/ras44/awesome-responsible-ai
Contributions and comments welcome.
If you're technical, feel free to submit a PR directly on GitHub. Or DM'ing me here works too.
I hope it's useful for those trying to get a quick idea of what this space looks like.
Thanks.
Saving for retirement might be unnecessary if AI Policy doesn't work out, so might want to focus there ;)
In the spirit of the OP's post, I propose a new post: does hiring Impact Analysts work? You might want to commission a study of impact analyst effectiveness in real-world situations that involve small and messy data with long durations to outcomes.
While an eager Impact Analyst might nerd out on all sorts of fun academic tools and theoretical proposals, in industry and at this stage of your growth, nobody should really care about measuring if your programs work and everyone should share about making them work as well as possible. Spoiler: they work. If you don't accept that, then you have to accept that your programs certainly aren't hurting anyone and shouldn't be prematurely derailed because someone is skeptical about whether or how well they work.
IMHO, define a short-term performance metric and put this budget toward a relentless and talented PM who is incentivized by the metric. They should be able to set up the basic data gathering (you don't need anything fancy), listen to customer feedback, and iterate on your product quickly.
Agreed here
There's only really no point in doing any work at all if you are confident the outcome is bad.
Thank you for sharing this. I largely agree with your views on how EA should evolve, but my main question is: if it did, would it still be EA and not just some kind of privately funded and significantly smaller version of the external ecosystem that already contains researchers, multilateral aid organizations, and governments? Is an evolved EA just a microcosm of what already exists elsewhere?
So what options does EA have?
Could it be that the EA brand (but not the community!) has become obsolete?
I would like to propose a related study:
Null Hypothesis: AI Safety talent programs are harmful.
Result: Null hypothesis rejected.
Study complete.
Not adding much here, but did the product team ever consider calling it "Witcher"?
Thank you for sharing @abergal
I find it encouraging that CG are investing in things like capacity building and relying on anecdotal evidence as insight. Ten years ago, my impression is this would have been almost EA-heretical in the face of "gold-standard" RCT canon. Reading this article, I'm more convinced that high-impact opportunities do exist to build and bridge AI Safety communities, and that it's a meaningful direction to contribute to. I can't help but think that there are unaddressed niches in uniting a relatively fractured ecosystem across industry, government, and research. Practitioners, policy-makers, and researchers are very important roles, but they are only part of the equation as anyone familiar with organizational behavior knows. Here's to the generalists who can also make many things happen :)
Congrats on the funding! And thanks for sharing so much detail on your vision and strategy. I've been very interested in contributing to this space and have been speaking with a variety of international university groups over the past few months about exactly these types of initiatives (at a more grass roots level). I'd be happy to help connect you with them!
Hi everyone, I'm Roland. I've worked in econometrics and experimentation for tech and consumer companies for many years and have been lurking on EA for the past couple. I'm a bit of a polymath, with a couple graduate degrees in computational physics and business administration. I'd like to engage more with people who are interested in working on effective and novel solutions to problems like extreme poverty, AI safety, and health. I am action-oriented and interested particularly in exploring low-friction, big-if-true POC solutions that can be trialed in these domains. I love meeting new people so feel free to say hi :)