Will Greg Brockman's bullish optimism on AI in 2024 prove true?
➕
Plus
71
Ṁ8391
Jan 1
20%
chance

Greg Brockman (President OpenAI): Prediction: 2024 will feel like a breakthrough year in terms of AI capability, safety, and general positivity about its potential impact. In the longer term, it'll look like just one more year on an exponential that can make everyone's lives better than anyone's today.

This resolves to YES is if, on 1 Jan 2025, this statement is generally believed to be true.

It resolves to NO if this statement proves is generally believed to be false.

If there is no market consensus and I am not confident in the winner, I will settle it via Twitter poll or other similar mechanism.

Get
Ṁ1,000
and
S1.00
Sort by:

How will this be resolved if part but not all of the statement is generally believed true (for example if there is a general consensus that AI has a breakthrough in capability but not safety)?

@AaronHahn presumably No, since (P ^ Q ^ R) is False if any of {P, Q, R} are False.