When will an AI-assisted bio-attack first result in at least 1,000,000 deaths?
41
248
1.5K
2046
1.6%
Now - 2025
5%
2025-2030
17%
2030-2035
6%
2035-2040
5%
2040-2045
66%
Never or after or including 2045

• The date will be the date of the release of the pathogen, not the date when it causes the deaths. This may require a delay in the resolution of the market. Similarly, establishing the cause of such a pandemic may require some delay to the resolution of the market.
• Market will be resolved according to official deaths (even though there are known issues with these figures)

• Whether an incident was AI-assisted or not will be resolved based upon on an official government investigation into the incident. Claims from a government that is unreliable will be disregarded.
• I just realised that I didn't actually manage to make the years unambiguous. So the period 2030-2035 includes 2030, but ends at the start of 2035. "Never or after 2045" actually means after the beginning of 2045.

Happy to further clarify the resolution criteria.

Get Ṁ200 play money
Sort by:

Seems hard to confirm in the most probable cases. What about gain-of-function research by a lab, aided by some machine learning technique (which is already pretty standard) followed by a leak? One would then need to ascribe intention to the leaker..

An accidental lab-leak wouldn't count as a bio-attack.

More related questions