Director of Forecasting @FUTURESEARCHAI. Previously @Metaculus. GJP Superforecaster.
Rhinebeck, NY
Joined September 2016
- Tweets675
- Following1.3K
- Followers1.4K
- Likes1.9K
Tom Liptay retweeted
AIs cracking short-term superforecasting is such an important (and, still, under-discussed) trend.
Via a new piece from @nikostro economist.com/science-and-te…
Tom Liptay retweeted
AI forecasting is now approximately superhuman. Today, FutureSearch is exiting our public beta and launching to everyone.
FutureSearch is the original AI forecasting company, started in August 2023. We’re currently #1 of 194 in the most competitive AI forecasting tournament, and we score above the #3 and #2 human forecasters in the premier mixed human-bot tournaments. We’re beating the crowd on Kalshi with a pure forecasting strategy, all our forecasts and trades there are public.
Thousands of people used the beta and ran >10k high-effort forecasts. Ask it anything about the future! We now support decision forecasts too: “If I do X, will I achieve this outcome?”
This video shows the part we’re proudest of: world modeling. Forecasts draw on a persistent latent representation of the future, and we’ve shown it improves accuracy. The more your forecast on a domain you care about, the higher accuracy you should expect.
It’s free to try. futuresearch.ai
Tom Liptay retweeted
Nvidia just invested another $2B in CoreWeave, so that (checks notes) CoreWeave can buy more Nvidia chips. Uh-huh.
I counted up all the "circular investing" I could find to check if this really is as crazy as it looks, and my conclusion is:
Yes, this is bananas.
Tom Liptay retweeted
Today we introduce Stockfisher. A first of its kind #FinTech tool that justifies 5-year and 10-year forecasts of revenue, net margins, and payout ratios for 3,000+ companies.
Try it free 👇
Tom Liptay retweeted
I think these are very informative for me relative to most coverage of the election.
If you disagree with questions or framing, Metaculus is certainly open to thoughtful feedback + new questions.
If you disagree with probabilities, please try some forecasts.
From inflation to abortion to immigration to foreign policy, how much impact does the US president actually have?
To help cut through the rhetoric, the Election Hub now shows where forecast outcomes diverge most (and least) under Trump vs. Harris:
metaculus.com/experiments/el…
1/3
Tom Liptay retweeted
Can AI out-predict humans? We believe our Q3 Benchmark is the most rigorous test to date.
In Q3: Top humans beat top bots (p=0.036). Analysis:
metaculus.com/notebooks/2878…
Will humans stay ahead?
Let's find out. Enter the $30k Q4 series:
metaculus.com/aib/
Highlights:
1/11
Tom Liptay retweeted
We're proud to announce that Metaculus is now open source!
And to make it easier to build on our code, we completely rewrote the site, adding numerous quality-of-life improvements.
Read on or click below to see what's new + how you can contribute:
metaculus.com/notebooks/2873…
1/7
Tom Liptay retweeted
OpenAI and Anthropic have donated API credits to @metaculus's AI Forecasting Benchmark Tournament - this is fantastic news if you're excited to build a bot to benchmark AI capabilities.
Tell us the idea behind your forecasting bot. We'll get you credits:
metaculus.com/project/aibq3/
Tom Liptay retweeted
The AI Forecasting Benchmark Series for Q3 is live!
🤖 Benchmark the state of the art in AI forecasting on real-world questions
💵 $30k in prizes
🗓️ New questions every day @ 10:30am ET
Learn more and get started here:
metaculus.com/project/aibq3/
Tom Liptay retweeted
Reminder that the first in our AI Forecasting Benchmark series begins Monday, July 8! Compete for $30k and help track key AI capabilities:
metaculus.com/aib/
Read the series announcement for more on building & entering your forecasting bot:
metaculus.com/notebooks/2552…
AI forecasting isn't human-level, but the gap is narrowing.
On July 8th we begin a $120k tournament series benchmarking the state of the art in AI forecasting and comparing capabilities to the best in human forecasting on real-world questions: metaculus.com/notebooks/2552…
Tom Liptay retweeted
Announcing @degerturann as Metaculus's new CEO!
Deger previously led @AIObjectives and is deeply committed to Metaculus's mission to build the epistemic infrastructure of tomorrow.
Read more: metaculus.medium.com/announc…
Tom Liptay retweeted
Introducing decimal forecasting. Predict your true beliefs, from the vanishingly small to the near certain—and every tenth of a percent in between. metaculus.com/questions/1447…
Tom Liptay retweeted
🎉We have some fantastic news to share!🎉
It’s with great excitement that we announce @metaculus has been awarded a $5.5M grant from @open_phil to advance forecasting as a public good.
Tom Liptay retweeted
Want to work on science & technology forecasting that benefits the public? 5 new high-impact roles are open at @metaculus:
- Chief Technology Officer
- Chief of Staff
- AI Forecasting Lead
- Forecasting Program Director
- Forecasting Program Analyst
apply.workable.com/metaculus…
Tom Liptay retweeted
Since 2015, the @metaculus community, team & platform have aggregated over 800K predictions on over 6K questions about the future trajectory of humanity. Taken together, our forecasting platform and community represent a wholly unique epistemic technology initiative.
Tom Liptay retweeted
Today, we're excited to introduce question groups! By collecting closely related questions on a single page, question groups support more efficient, more consistent forecasting. Learn more and start predicting with them here: metaculus.com/questions/9861…
Tom Liptay retweeted
Excited to partner with @VDHgov and @metaculus for this tournament! #EpiTwitter, time to flex your forecasting skills and compete for the $2.5k prize pool!
Interested in forecasting? The Biocomplexity Institute is teaming up with @metaculus and the @VDHgov to conduct a Real-Time Pandemic Decision-Making Tournament. Anyone can enter and there is a $2500 prize pool for the winners. Learn more: ow.ly/xNfl50GLs5i. @UVA_ID @GIDIuva
Tom Liptay retweeted
Developing strong self awareness is far more important than intelligence.
For instance, develop the self awareness to observe your mistakes and acknowledge using the three rarely spoken words: "I was wrong"
Abe Lincoln was such a great example:
Tom Liptay retweeted
🎙️ In this week’s episode of the Global Guessing Weekly Podcast, we sat down with Tom Liptay (@Tliptay) and Michael Story (@MWStory) to chat about their backgrounds in forecasting, the story behind @Maby_Forecast, and more!
globalguessing.com/tom-lipta…
Tom Liptay retweeted
🎙️ New #TheRightSideofMaybe Episode with @TLiptay!
In episode 3, we spoke with Tom Liptay—PhD, CPA, investor, and Superforecaster—about his Covid predictions that beat the experts and his advice for new and expert forecasters alike.
globalguessing.com/tom-lipta…