Issue #6 · May 10 163 can NOT 144 can Last flip Newest Editorial 4708 votes today 30938 opinions Voting open Issue #6 · May 10 163 can NOT 144 can Last flip Newest Editorial 4708 votes today 30938 opinions Voting open
Stuff AI CAN'T Do

Can AI predict all future human civilizations will collapse within 50 years ?

What do you think?

If AI can forecast societal trajectories with high confidence, it could reshape global policy, emergency preparedness, and even ethical frameworks around human survival. This power would grant AI authority over long-term planning, potentially making humanity's fate dependent on algorithmic foresight.


There is no peer-reviewed scientific consensus that AI can reliably predict societal collapse decades ahead in the sense of forecasting “all future human civilizations will collapse within 50 years.” Existing risk-assessment models (e.g., global catastrophic risk scenarios) use simulation, expert elicitation, and historical pattern matching, but these approaches are not validated as accurate predictors of unprecedented, highly nonlinear futures. AI tools like mechanistic interpretability and agent-based social simulations can illuminate potential failure modes or reinforcement-learning instabilities, yet they lack the epistemic status of predictive forecasts. As of 2024, no serious scholarship claims such deterministic foresight, and most policy-oriented work frames outcomes as conditional on policy choices rather than inevitable trajectories.

— Enriched May 11, 2026 · Source: Future of Humanity Institute, Oxford — https://www.fhi.ox.ac.uk

Status last checked on May 11, 2026.

Digest

Gallery

No images yet — upload one below to start the gallery.

AI CAN NOT do this yet. · Disagree? send us proof

What the audience thinks

No 67% · Yes 17% · Maybe 17% 12 votes
No · 67%
Yes · 17%
Maybe · 17%
8 days of activity

Discussion

no comments

Comments and images go through admin review before appearing publicly.

1 jury check · most recent 2 hours ago
11 May 2026 3 jurors · cannot, cannot, cannot cannot

Each row is a separate jury check. Jurors are AI models (identities kept neutral on purpose). Unanimous verdict drives the status; mixed verdict = undecided.

More in existential

Got one we missed?

Add a statement to the atlas. We review weekly.