Notes on Manifest

A festival for forecasting and prediction markets, hosted by Manifold and Manifund.

Last week I attended Manifest hosted by the Manifold and Manifund teams. This is a conference that was dedicated to forecasting and prediction markets. Their shirts had the slogan “predictions, markets, mechanisms”. In reality is was more like 1/3 prediction markets, 1/3 AI (mostly safety), 1/3 rationalist/EA/other.

This was the best conference festival I have attended (and a very stark difference from crypto conferences).

It was hosted at the Lighthaven space with about 500 people, tons of “soft seating”, interesting talks, and tons of engaged attendees.

I thought it was such an interesting event here are some random notes I had on it.

Notes

• Many of the talks and events were actually attendee run. They had this custom scheduling app in which there were a good amount of headline speakers (like Byrne Hobart, Scott Alexander, Nate Silver, Patrick McKenzie). But then they also let attendees simply grab open timeslots and host whatever they wanted (these events were sneakily good)

• People care more about knowing what is right rather than being right or making money. This is more of a comment with regards to crypto conferences. It was refreshing!

• There was this talk about election contracts, the recent CFTC ban proposal, and Kalshi suing the CFTC. Luana from Kalshi and the team from Predicit were in the audience and basically commented they are both well funded and looking to take this as far as needed (Supreme Court etc.)

• Interesting talk on political prediction markets from star spangled gambler podcast

  • Discussion on how kicking out Richard Hanania for offensive comments in various forecasting communities was wrong (there was a good amount of debate on this) and how maybe the forecasting community should embrace more troll-ey comments like the crypto community. This gets me thinking, will EA community kneecap the growth of these markets simply because they decide something (that is actually driving growth) isn’t moral or just or rational or something.

    • Kinda funny enough in the “hard question for Manifold team” talk one of the questions was should Manifold take EA grant funding or not (they have in the past). As in like is it a good use of EA grants to fund prediction markets or should they be focusing on things like AI Safety and whatever else is trending on LessWrong.

• Nascent a crypto VC is hiring for Sovereign a prop fund with key attributes about people that trade prediction markets. You can check out the form here, I think this is pretty cool and for sure ups my ranking of Nascent for top crypto VCs.

• Nate Silver is becoming a advisor to a prediction market company (announcing it soon).

• Scott Alexander mentioned this, “I think that prediction markets are very helpful to get acquainted with a subject, no need necessarily for the rationale. Just wanna know what people are thinking”. I find this interesting because a lot of the talks were about how to provide evidence for a forecast, etc. But it seems like just simply providing a percentage chance of something next to a news article might be helpful.

• There was a startup pitch competition. All the startups are documented here. This was cool because there was also live betting on if a startup was going to be funded.

  • One of the pitches was $2m for prediction market on Urbit. Feels like it’s gonna fall into the same thing as Vega. They are shipping libraries and packages and things but unsure if they will actually get out a product that people use.

• Future Search (ex google and Metaculus guys) are pushing LLMs to their limits and using them to aid the argerious task of research in forecasting. They had a pretty impressive presentation that was basically a walk through on how to predict this market

• SFCompute had this talk called “Design Your Own Compute Market”, which was very interesting. It was a collaborative workshop with a few prompts and ideas which were aggregated into this document. It ended with the team hinting that they are launching a futures market on GPU compute in the summer (August)!

Overall Manifest was a great conference. I hope they have it again next year.

I am launching a prediction market data platform soon! It is featured alongside a robust RSS news reader so you can discover related markets while you read.

Join the waitlist at adj.news