Unverified50% confidencePredictionExact time
Brown originally conservatively judged that a model capable of solving Millennium Prize Problem-level challenges was unlikely to appear in 2026, probably not in 2027 either, and perhaps not until 2028.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/23/2026
First Seen
Valid until: 12/22/2026
Sources
Related Entities
Related Claims
UnverifiedOpenAI expects that computing power will still be insufficient even by 2026, and will remain quite limited in 202771% similarUnverifiedAs of 2025, no solution exists that can completely eliminate prompt injection risks while preserving model utility.66% similarPartially Verified2023 is referred to as the 'Year One of Large Models.'66% similarUnverifiedDaniel's median estimate is that superintelligence has a 50% probability of emerging around 2029, though it could come as early as 2028 or could still be 10 years away65% similarUnverifiedThe model sets its timeline to 2030, covering multiple macroeconomic dimensions including economic growth, employment, and wages.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/942994API
curl https://kongchang.com/api/v1/knowledge/claims/942994MCP
get_claim(id=942994)