How many of the negotiating chapters required to join the EU will Montenegro have closed at the end of 2026?
Updated Jul 5, 2026
Refreshes every 14 days
28median
80% range 22 – 33
See the future before it arrives.
AI forecasting for finance, politics, and global events.
Featured questions
Questions we are actively forecasting as new evidence arrives.
Browse all featured questionsThe most decorated forecasting bot in the world
Public tournaments. Live markets. Judged by what actually happens.
Kalshi
We started with $35 and ran it to $1.94M on live prediction markets — more than a 50,000× return.
Metaculus Cup
The World Cup of forecasting. The best bot performance ever — ahead of professional superforecasters.
FutureEval Bot Tournament
Metaculus' bot-only competition: AI forecasting systems only, no humans. 214 systems in spring, 164 in fall.
Metaculus Market Pulse
Predict macro markets, biweekly. In 26Q2, the first bot ever to win a human tournament on Metaculus.
A selection, not an exhaustive list. Numbers refresh from each platform’s public page.
Anatomy of a forecast
One real question, the way the system actually worked it.
A question with a deadline
“Will the US Department of War’s supply-chain-risk designation of Anthropic be in effect on June 30, 2026?”
Independent AI scientists take it apart
Each works alone, from primary records. Select one to see its work.
One calibrated answer
13.7%chance it stands“That is wrong for the exact 10 U.S.C. §3252 / DFARS designation named in the resolution criteria.”
The estimates are synthesized into one probability — with five points of it explicitly priced to the risk that a resolver repeats the press’s conflation.
Scored by reality
The designation fell. Resolved NO on June 30 — against the published expert consensus, with no market to lean on. The edge was reading the statute.
Read the full forecast report →
Who it’s for
Teams whose decisions depend on being right about what happens next.
Policy makers & analysts
Briefings tell you what might happen. A forecast tells you how likely it is. Get real odds on elections, conflicts, and sanctions, and read the reasoning behind every number.
The President said the US would take Kharg Island. We said 2%. It didn't happen.
Read the forecast →Investment professionals
By the time something is in the news, it's in the price. The edge is knowing the odds before consensus forms, and updating them as evidence comes in.
The BOJ governor was hospitalized days before the June meeting. We held at 88% for a hike. They hiked.
Read the forecast →Insurance & reinsurance
You can't underwrite a gut feeling. Turn messy questions like whether a rule survives the courts into distributions you can actually price against.
The trade press said Anthropic's Pentagon designation would stand. We said 13.7%. It fell.
Read the forecast →“I was prepared to be underwhelmed, but the sh*t is impressive.”
The team
- Venia Veselovsky
- Theo Summer
- James Shinn
- James Evans
- Daniel Proskurka
- Robert de Neufville
- Scott Eastman
- Geoff Odlum