Will I consider myself a full-time AI alignment researcher by the end of 2023?
17
168
430
resolved Jan 1
Resolved
N/A

Please use all available information about me to make this prediction. I will also answer questions in the comments.

Get Ṁ200 play money
Sort by:
predicted YES

This market was worded in a confusing way, and my comments attempting to clarify the market were also confusing.

I think it is genuinely ambiguous how this market resolves.

Sources of ambiguity:

  • Definition of “full time”

  • Definition of “by the end” for a conditional that can flip many times over a year

  • Definition of “alignment” given my growing interest in adjacent AI safety subjects.

Thus, N/A.

predicted YES

I’m currently funded to do AI alignment research. However, the funding comes in the form of an up-skilling grant. It seems ok to resolve this market to yes, but let me know if you have objections.

@Alana Are you currently producing AI alignment research full time, or upskilling?

predicted YES

@ThomasKwa The line’s getting blurry. I think it’s still mostly up-skilling right now, but this might change.

sold Ṁ64 of YES

Liquidating across manifold

predicted YES

I'm going to sell some YES shares and use the money to subsidize the market.

sold Ṁ25 of YES

@AlanaXiang Bought back :)

bought Ṁ10 of YES

Sold back to 84 for now. Want some extra liquidity.

bought Ṁ10 of YES

@AlanaXiang Bought back to 85 :)

predicted YES

I have subsidized this market with M50 of additional liquidity.

Are you trying to become a full-time AI alignment researcher? What's stopping you? (Or are you already one?)

predicted YES

@KatjaGrace

Yes, I am trying to become a full-time AI alignment researcher. I currently work part-time at ARC (evals).

A few scenarios where this market may resolve positive:

  • I do SERI MATS and then get an alignment job

  • I directly get an alignment job

  • I receive funding for independent alignment research

  • I save enough money to do independent alignment research without funding. In this case, I'd probably only consider myself a full-time alignment researcher if I feel like I am making research progress.

A few scenarios where this market may resolve negative:

  • I change my mind about AI alignment research being my top priority. Instead, I may decide to focus on evaluations, forecasting, or biosafety

  • I fail to get a job or funding for alignment research

  • I decide I want to stay in college vs leaving to work on alignment full-time

  • I decide I want to focus on up-skilling instead of full-time research

  • I do full-time alignment research before the end of 2024, but for some reason decide to stop before the end of 2024.

bought Ṁ0 of YES

@AlanaXiang Note that if I am doing SERI MATS or a similar mentorship program by the end of 2024, then this market would still resolve positive. However, if I am doing some other AI alignment training program that did not have a major research component, then this market would resolve to negative.

predicted YES

@AlanaXiang The dates here don’t match the title. Please defer to the title.