COO at Redwood Research (applied AI alignment). Formerly an early software engineer at Aurora, a self-driving car startup.
I think $20B in 4y is somewhat of an outlier, even among super successful billionaire founders. Eg a few quick googles (assuming CEO has something like 10% of the company after several rounds of dilution)
I agree that selection bias and survivorship bias affect things like this, and it probably would’ve been good to call those out explicitly. I have a draft part 2 of this post that discusses that and how hard it is to get a job at one of these companies.
I also appreciate the comment directed at people that might feel alienated by the post, and agree that I don’t want those people to feel alienated by this. A more positive frame on this post would be: some people think that they can only work at the ~10 EA orgs they think most highly of to have a good next step, and I think there are at least ~100 startups that they could also consider that would also be very promising. I think in general, it’s easier for an early career EA to get a job at a fast growing startup than one of the top EA orgs.
I disagree with the point about FAANG being a better option. I agree it’s a solid, lower variance option, but I think it’s higher expected value to try to get a job at a fast-growing startup for people that feel like they could do that.
But anyways, I’m sorry to anyone that felt alienated by this post, and I think not feeling like you can get this kind of job doesn’t mean that you can’t do lots of good things for the world.
I have a draft part 2: it’s easier than it sounds. One of the reasons I believe that is because many of the best-of-the-best startups have many vc’s that want to fund them (so it’s not as hard for them to identify which are the best, but it is hard for them to compete to be the one to fund it). On the other hand, these startups need all the excellent employees they can get.
Makes sense, upvoted. I like “fast-growing” more than “top,” because “top” makes me think more “is already Airbnb” vs “could be the next Airbnb.” Maybe the best term would be “exceptionally fast-growing” or “exceptionally likely to be successful.”
Fast and successful are definitely a spectrum, and it seems definitely true that somewhat successful is still good for career development. I think one claim I didn’t spell out fully is that people aren’t selective enough in choosing which startups to work at. Of friends who worked at companies in the “YC or Techstars startup” reference class, I think several had really positive experiences, but several worked at companies that went under and weren’t that positive, and it seems to make a big difference to choose one that is exceptionally good vs ok.
We're a nonprofit. We don't have plans to make profits, and it seems less likely than e.g. OpenAI that in the future we would go nonprofit --> tandem for-profit / nonprofit, but there are a variety of revenue-generating things I can imagine us doing (e.g. consulting with industry labs to help them align their models).
(As of ~4/1, I've seen a much higher anecdotal success rate w/ strategies like this)
In hindsight, I would've liked to spend more time editing my original post (I was trying to get it out quickly). I think the framing I would use in hindsight is:
I'm hugely in favor of young healthy people finding vaccines that are either 1) going to be thrown out, 2) at a place w/ extra time slots that wouldn't go to someone else, but 3) from the evidence I've seen, it seems like 1/2 aren't options right now, and so 4) trying to find a way to get a vaccine right now seems likely to be taking directly away from someone else who is higher risk. There are still other ways to justify this (e.g. catherio's comment about ethical offsetting), but because I've heard a bunch of rumors along the lines of "1/2 are options right now," I wanted to share that the evidence I've seen suggests that's largely false. It's probably also not a big deal if someone gets a vaccine early, as it seems likely people will be able to seek them out within the next ~month.
I agree that slow rollout is more of a problem than poor prioritization. But at least in the Bay Area, I think trying to get a vaccine as a young healthy person right now is in expectation directly taking away a vaccine from a more at-risk person for x weeks, and haven't seen significant evidence against that yet. As soon as there is the ability to e.g. drive somewhere to get vaccinated where the time slot would have otherwise gone to waste, I'm strongly in favor of doing that.
Agreed that driving somewhere seems like a good option -- that's what I was trying to allude to with my comments like "there appear to be some mass vaccination events that people are likely to be able to access in the next month." That being said, I know of many people who've tried to do exactly that and have not found available appointments. I mentioned in the comments below that the website linked there for instance didn't have any options within ~1hr of the bay, although I found one 3.5 hours away.
Ya that seems like a decent approach for people that want to get one early. To be clear, I know of people who are making the arguments I said above to justify getting one early (and who have no plans to offset), and I think there are other arguments to justify getting it early, but I don't think "there are many opening available in the bay" is backed up by evidence from what I can tell.