Skip to main content
MANIFOLD
Will there be serious AI safety drama at Meta AI before 2026?
43
Ṁ1kṀ6.5k
resolved Mar 3
Resolved
NO

AI safety meaning AI not killeveryoneism, not racism/bad words etc.

The drama has to be at least somewhat public facing.

The qualification of 'serious' will be in my judgment. I won't place any bets.

Market context
Get
Ṁ1,000
to start trading!

🏅 Top traders

#TraderTotal profit
1Ṁ333
2Ṁ121
3Ṁ89
4Ṁ86
5Ṁ67
Sort by:

@SemioticRivalry Please resolve

I have a feeling this happened but it wasn't made public

if this market concerned openai and was created before the altman firing saga, would you have resolved it YES due to the altman drama?

Probably meaningless posturing ahead of the Summit in France, but fwiw: https://techcrunch.com/2025/02/03/meta-says-it-may-stop-development-of-ai-systems-it-deems-too-risky/

What if it's not bad words or notkilleveryoneism but something intermediate like chemical weapons?

Base rate is low, as far as I remember none ever happened at Microsoft, Amazon, Apple or Google (nor at Google Brain, Google Deepmind). OpenAI seems to attract this to a large degree, for some reason, but it's a newer company.

@SemioticRivalry Does the drama need to be internal? Like alternatively it could be that some external entity tell meta folks to do AI safety a certain way and they object publicly in a dramaful way

@Bayesian preferably internal, or if external it would have to be a significant news story, like would my parents have heard about it