Hmm, this is interesting. I think the ethical concerns below are mostly irrelevant. Both X-risk and the pedestrian OpenAI fraud and dishonesty are public knowledge and have been for a while, so anyone who's going to work there is aware of both and has decided this is not enough of a downside to outweigh all the money they'd be making. In general it seems quite unlikely to me that anyone with such a personality would happen to change their mind on the ethical aspects in a few years; that sort of core personality reversal generally requires something dramatic happening to the individual or a close friend or family member.

However, the creator says they're a senior, and people at that age are still pretty malleable. e.g. I stole some money from a friend when I was 14 or so, but grew out of the behavior and apologized+returned it a few years later. So if Medford is 18 currently, I could see them still having a youthful selfishness at present but growing a conscience by 25 or whatever.

(There's also the fact that they're posting about this on Manifold, which implies more immersion in the X-risk community than the average OpenAI employee. But again that would already be factored in to their decision to accept the offer.)

So I think this market is actually mostly about whether OpenAI ends up doing poorly as a business, perhaps losing to Google or Anthropic. Since Medford seems to be profit-motivated, my guess is that their regret will be judged by whether they could have made more money elsewhere.

Is there a significant chance that you'll come to believe OpenAI is endangering humanity, and therefore regret helping them in that?

@AhronMaline I don't believe this now and I think it's unlikely I would by the resolution date, unless I learned something nonpublic that changed my beliefs. However, if I came to believe this, I would regret working there.

@Medford what is your opinion on these articles?

• https://philosophybear.substack.com/p/hydra

• https://philosophybear.substack.com/p/we-need-to-do-something-about-ai

In general do you normally regret decisions? Or mostly think the grass is greener on the other side?

@kamn I put some examples of decisions I did and didn't regret in the description. For a few more:
- I spent a lot of time in late HS focused on college applications even though I found this uninteresting and it didn't feel aligned with my values. I did not regret this.
- I chose majors that I later switched out of, twice. I didn't regret this either time, since I made the best decision I could have with the information I had pre taking classes in the major.
- I didn't read much in high school, and I regretted this.
- I regretted not taking more math and science classes younger, and not making a more active effort to truly learn in the ones I took.
- I regretted doing poorly in a class in HS (the poor decision being not to try harder in the class), but I have never regretted doing poorly in a class in college.

I would guess that I'm in maybe the 60th percentile as far as regretting things (with higher percentiles having regrets more often)? I know plenty of people who regret stuff more often, and plenty (slightly more) people who regret stuff less often.

For context, I took an internship at a big tech company instead of a smaller company and regretted it, since I did not learn very much. The previous summer I interned at a smaller tech company and did not regret it. The following summer, I interned at a non-tech company and didn't regret it, but the amount I learned was partially due to the novelty of the work which might wear off after more than a summer.

@Medford Also, I am not very concerned with earnings shortly after graduating, so if I made a lot of money (for a new grad) but felt like I was pigeonholed or didn't learn, I would likely regret the job.

@Medford Also, this is implied by "coder" in the title but to clarify, I would not be working on research, I would be a software engineer.