“Few things are either certain or impossible.” - Phil Tetlock & Dan Gardner in Superforecasting. goodjudgment.com/services/futur…
English
Good Judgment Inc
2.6K posts
Good Judgment Inc
@superforecaster
Delivering insights into global challenges through the science and art of #Superforecasting. Test your forecasting prowess at https://t.co/5H8XIhN7eK
Katılım Eylül 2014
698 Takip Edilen11.3K Takipçiler
“Human beings have coped with uncertainty for as long as we have been recognizably human.” - Phil Tetlock & Dan Gardner in Superforecasting.
goodjudgment.com/services/futur…
English
“Ultimately, it’s not the crunching power that counts. It’s how you use it.” - Phil Tetlock & Dan Gardner in Superforecasting. goodjudgment.com/services/train…
English
Link to post: goodjudgment.substack.com/p/let-me-get-t…
English
Millions of dollars trading on a Polymarket contract where “severely restricts” wasn’t defined until six weeks after launch. “Making up standards as you go does a disservice to market participants.” See Ryan Adler’s “Let Me Get This Strait” 👇
English
Polymarket bundled two tariff questions into one. The Supreme Court happened to treat them the same way, but it easily could have split the baby, as courts often do. Good question design matters, writes Ryan Adler, Superforecaster and Good Judgment Question Team Lead. Link👇
English
“Team members must disagree without being disagreeable.”
- Phil Tetlock & Dan Gardner in Superforecasting.
goodjudgment.com/services/upcom…
English
Our highlights from the full report: goodjudgment.substack.com/p/what-superfo… 2/2
English
FRI’s latest survey found Superforecasters “bullish” on an AI forecasting benchmark. What the blog post didn’t mention: they also said hitting that mark wouldn’t mean much. “Skill lies in knowing when AI number crunching will be enough. Judgment, in knowing when it won’t.” 1/2
English
Good Judgment Inc retweetledi
@shashj @Noahpinion Superforecasters actually agree that AI will outperform their 2024 benchmark in 2 years. ForecastBench has a broad scope and is not fully representative of the dominant edge that humans still hold in questions where data is sparse and judgment is required x.com/i/status/20259…
Forecasting Research Institute@Research_FRI
🔮 AI systems are expected to surpass top human forecasters within the next few years Superforecasters predict that AI systems will outperform superforecasters on ForecastBench, a forecasting benchmark, by 2028. That's faster than both the median expert (2030) and public (2033) forecast. Forecasters disagree on the significance of this milestone—some note that AI excels at data-rich, quantitative questions but struggles with geopolitical questions where data is sparse.
English
What you’ll practice: debiasing, teaming for accuracy, and framing rigorous questions that drive decisions. 10 & 11 March, 12:00 to 2:30 pm ET. Details: goodjudgment.com/services/upcom… Grab a seat: tickettailor.com/events/goodjud… #Superforecasting
English
New from @Research_FRI: a wet-lab RCT tested whether LLM access helps novices execute real bio tasks. Result: no uplift.
Small sample, one study. But Superforecasters outperformed domain experts at predicting the outcome. The edge: base-rate discipline.
Calibration matters.
Forecasting Research Institute@Research_FRI
In mid-2025, @ActiveSiteBio told us about an RCT it was working on that would test how access to LLMs affected novices' ability to do molecular biology work in a wet lab. We ran a forecasting study alongside this RCT, to find out what forecasters predicted for the RCT outcome and to understand how the study results connected to real world biorisk. So, how well did superforecasters and experts predict wet lab skill uplift from LLMs? 🧵 Read on to find out…
English
@stevehou @justinaknope @mansourtarek_ Good Judgment has been identifying and fielding Superforecasters for over a decade. The NYT piece from Feb 11 covers how that approach compares to prediction markets and AI: nytimes.com/2026/02/11/bus… (may be paywalled)
English
@justinaknope I think one interesting line of research would be clearly to leverage the Kalshi data to identify such a team of “superforecasters”, which I think @mansourtarek_ had mentioned was one of the books that inspired him as well.
English
On the same topic, now Good Judgment has a blog post out saying their "superforecasters" are more accurate at predicting central bank decisions than Polymarket
goodjudgment.substack.com/p/testing-poly…
Justina Lee@justinaknope
We wrote about the paper on Kalshi's macro markets last month but I see it's gaining traction now One caveat: Kalshi beat fed funds rate on calling Fed decisions because of ONE single meeting: Sept 2024, when the Fed surprised with a 50bps cut, which the Kalshi market put a higher weight on
English
2/2 Substack (short version): goodjudgment.substack.com/p/meet-superfo…
Blog (full interview): goodjudgment.com/about/our-team…
English
Meet a Superforecaster: AI risk management expert Malcolm Murray explains why AI experts often overestimate specific risks and adoption timelines, how he uses Superforecasting best practices to assess biological and cyber risk, why resilience is the path forward, and more. 1/2
English
“The point of making forecasts is not to tick all the boxes on the ‘how to make forecasts’ checklist. It is to foresee what’s coming.” - Phil Tetlock & Dan Gardner in Superforecasting.
goodjudgment.com/services/upcom…
English
“In AI, there’s no point in making random predictions. But there is still a great need for detailed and elaborate forecasts.” Meet AI risk management expert & Superforecaster Malcolm Murray.
goodjudgment.com/about/our-team…
English
“Learning to forecast requires trying to forecast. Reading books on forecasting is no substitute for the experience of the real thing.” -Phil Tetlock & Dan Gardner. Join the March workshop (10 & 11). Details: goodjudgment.com/services/upcom… Grab a seat: tickettailor.com/events/goodjud…
English
Good Judgment Inc retweetledi
Good Judgment is featured in today’s @nytimes on the future of forecasting. Our CEO Dr. Warren Hatch @wfrhatch discusses how Superforecasters have an edge when the data is sparse and the environment is in flux.
nytimes.com/2026/02/11/bus…
English