PortfolioLeaderboard

 

TermsPrivacySecurityBusinessProtocolFeesStatusContact

Problee Money is for play. It can't be cashed out or moved.

© 2026 Problee, Inc.

Recorded on Base

AllElectionsCryptoBusinessAITechMovies & TVWorldPoliticsFavorites

Next verdict

BTC above $83,650.00 on Sep 30?23:59 UTC

Biggest moves today

  • +55%Will an AI coding agent score 62+ on the Artificial Analysis Coding Agent Index averaging 14.7 minutes or less per task by December 17, 2026?
  • +49%WLD above $0.4836 on Sep 30?
  • +43%BNB above $759.00 on Sep 30?

Elections

  • Democrats win Nevada governor?36dYes50%
  • Democrats win Georgia Senate election?36dYes95%
  • Democrats win New Hampshire Senate election?36dYes85%
  • Democrats win Michigan Senate election?36dYes71%

Elections

  • Ken Sim wins Vancouver mayor?19dYes35%
  • Pete Fry wins Vancouver mayor?19dYes66%
  • Kareem Allam wins Vancouver mayor?19dYes2%
  • Brad Bradford wins Toronto mayor?28dYes12%

Closing soon

  • Meta's Watermelon released by September 30, 2026?23:59 UTCYes2%
  • US x Iran ceasefire continues through September 30?23:59 UTCYes99%
  • BTC above $83,650.00 on Sep 30?23:59 UTCYes50%
  • SOL above $119.00 on Sep 30?23:59 UTCYes26%
  • XRP above $1.490 on Sep 30?23:59 UTCYes53%
All 144 open markets

 

 

 

 

Back
Closes in 78d

Will METR publish a study showing AI makes experienced developers significantly faster by December 17, 2026?

OPEN · Closes Dec 17, 2026, 23:59 UTC

Opening price · no trades yet

How a market resolves
  1. 1

    Trade

    Buy and sell shares while the market is open. Prices show what the crowd expects.

  2. 2

    Close

    Trading stops and the creator posts the result.

  3. 3

    Result posted

    The posted result can be challenged before it becomes final. A challenge sends it to a community vote.

  4. 4

    Community vote

    Members stake PM to vote on the correct outcome. Call it right and you earn the stake back.

  5. 5

    Settled

    The result is finalised and every winning share pays out. Anything left unresolved is reviewed first.

Full resolution rules

METR's famous 2025 study found AI made experienced open-source developers 19% slower, and its follow-up was inconclusive. Will METR publish rigorous evidence of a real speedup before year end?

Price History

PriceSharesTotal (PM)
Spread: 0.026Mid 0.02
Sign in to comment

—available
If correct
--
Average price
--
Fee
--

Resolution rules

How this market resolves

  • 1
    Resolution criteria

    Resolves YES if, after September 18, 2026 and at or before the close, METR publishes a study that randomly assigns experienced software developers to work with or without AI tools on real software tasks, and whose main task-completion-time result shows a speedup with a 95% confidence interval that excludes no effect. Self-reported productivity, non-randomized comparisons, benchmark-only experiments, and reposts of earlier results do not count. If no qualifying study is published by the close, this resolves NO. Source: METR (authoritative).

Related Markets

Will an AI score 80%+ on ARC-AGI-3 with the Standard harness for $26,098 or less by December 17, 2026?

Will an AI model score 53+ on the Artificial Analysis Intelligence Index at $1.63 or less per task by December 17, 2026?

Will an AI coding agent score 62+ on the Artificial Analysis Coding Agent Index averaging 14.7 minutes or less per task by December 17, 2026?

Will METR report an AI with an 8-hour time horizon at 80% success by December 17, 2026?