Problem: the term "AGI" was already ambiguous, and suffered of goalpost moving. I'm betting some YES due to the view, held by some, that GPT is already AGI.

Given the original tweet got deleted, here's a screenshot of a screenshot from the account, for posterity:

Q*

optimal policy achieved?

https://www.theinformation.com/articles/openai-made-an-ai-breakthrough-before-altman-firing-stoking-excitement-and-concern

YES traders, what company do you think believes they already have AGI? OpenAI? Google?

https://www.reddit.com/r/singularity/comments/16sdu6w/rip_jimmy_apples/k2aroaw/

I’m betting NO based on these two vague judgements:

1. AGI feels like fantasy to me. That’s not a reasoned opinion; that’s how I feel. However, I also think that it makes a lot of sense to judge AGI as unlikely to have been developed at this early stage.

2. Although I do agree with the markets that it’s more likely that an AI lab thinks they’ve achieved AGI than that they actually have, I have the vague idea that people at AI labs are “smart” and that they wouldn’t have such an erroneous belief. This is shakier than 1, and I understand people betting this market up on the grounds of human fallibility.

Any comments/ things I haven’t considered?

Resolves based on the views of lab leadership presumably?

@MartinRandall That'd work, whatever would make it into an official stance of the organization.

https://twitter.com/sama/status/1705752292484624863 hmmm

@dmayhem93