Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The AI Superforecasters Are Here

astralcodexten.com

Post reply on HN
Print view View original

21–30 of 60 posts

Re: The AI Superforecasters Are Here

#21

Post by citizenpaul »

datadrivenangel wrote: post #2

If the forecasting models were so good that people were actually consistently beating prediction markets, they wouldn't be starting startups to be selling it. And even if it is good enough, once you're shelling out thousands of dollars a year in research costs, does that give you any remaining alpha?

The market of trading is smaller than the market of business. I don't think most people have ever actually tried trading for real. The frequency with which you can put in a buy/sell order and nothing happens. Then you raise/lower the price several times and nothing still happens because nobody wants to buy or sell at a price that makes you money is much higher than you would imagine. If you've never tried it yourself.

That said, the very concept of selling such information means that it would eliminate any edge and become zero profit anyway.

Re: The AI Superforecasters Are Here

#22

Post by fer »

datadrivenangel wrote: post #2

If the forecasting models were so good that people were actually consistently beating prediction markets, they wouldn't be starting startups to be selling it. And even if it is good enough, once you're shelling out thousands of dollars a year in research costs, does that give you any remaining alpha?

Agree, to an extent. The link certainly smells like an ad. If the predictions are on lower volume markets (i.e. no institutional investors with actual quants behind) the compounding power stops. Also that means it can't beat institutional investors. So the reasonable option is to monetize what you can.

Re: The AI Superforecasters Are Here

#23

Post by davedx »

citizenpaul wrote: post #21
datadrivenangel wrote: post #2

If the forecasting models were so good that people were actually consistently beating prediction markets, they wouldn't be starting startups to be selling it. And even if it is good enough, once you're shelling out thousands of dollars a year in research costs, does that give you any remaining alpha?

The market of trading is smaller than the market of business. I don't think most people have ever actually tried trading for real. The frequency with which you can put in a buy/sell order and nothing happens. Then you raise/lower the price several times and nothing still happens because nobody wants to buy or sell at a price that makes you money is much higher than you would imagine. If you've never tried it yourself…

That only happens if you're trading very illiquid stocks.

Re: The AI Superforecasters Are Here

#24

Post by povik »

datadrivenangel wrote: post #2

If the forecasting models were so good that people were actually consistently beating prediction markets, they wouldn't be starting startups to be selling it. And even if it is good enough, once you're shelling out thousands of dollars a year in research costs, does that give you any remaining alpha?

If a model demonstrates being good by winning on prediction markets I may want to ask the same model questions which are too niche to have markets. That said I don't know if niche questions are enough of a revenue stream to be interesting for model providers.

Perhaps they may enforce a knowledge cut-off for information retrieval and price the service based on how recent the cut-off is, and also use the cut-off as a way to guard their advantage on the markets.

Re: The AI Superforecasters Are Here

#25

Post by bee_rider »

datadrivenangel wrote: post #15
energy123 wrote: post #12

Earlier quoted context omitted.

They do. One guy with at least 7 figure PnL is a well known forecaster, SemioticRivalry.

Only news article I found on that suggests the guy is mostly just market making against bets he thinks are dumb? No mention of AI.

Wait, the original blog post was about competing with top-tier guessers using “AI.” I wonder if there’s more money to be made on the flip side: find the people making dumb bets, at scale, and take the free money they are offering, haha.

Re: The AI Superforecasters Are Here

#26

Post by citizenpaul »

davedx wrote: post #23
citizenpaul wrote: post #21

Earlier quoted context omitted.

The market of trading is smaller than the market of business. I don't think most people have ever actually tried trading for real. The frequency with which you can put in a buy/sell order and nothing happens. Then you raise/lower the price several times and nothing still happens because nobody wants to buy or sell at a price that makes you money is much higher than you would imagine. If you've never tried it yourself…

That only happens if you're trading very illiquid stocks.

I can't tell you the number of times of put in an order on SPY,QQQ,SPX when I found favorable statistical analysis and I could not get filled at any price below the estimated profit potential. Those are some of the most liquid products in the word. I'm not talking pennies either.

Futures are the only thing for me this has never happened on. However no retail broker I'm aware of allows trading futures though API.

Re: The AI Superforecasters Are Here

#27

Post by janalsncm »

datadrivenangel wrote: post #2

If the forecasting models were so good that people were actually consistently beating prediction markets, they wouldn't be starting startups to be selling it. And even if it is good enough, once you're shelling out thousands of dollars a year in research costs, does that give you any remaining alpha?

The reason is, customer might care a lot more about the prediction than you do, or anyone else does.

For example, your customer might really care a lot about some niche prediction like the number of car break-ins in Walmart parking lots. In practice you won’t have sufficient liquidity in a prediction market to actually profit off of that prediction. But a security company might really want to know the answer to it.

Re: The AI Superforecasters Are Here

#28

Post by jdmoreira »

guelo wrote: post #11
AIorNot wrote: post #7

For Pete’s sake lets outlaw prediction markets already instead of hyping them in crappy greed-posts like this one I am embarrassed this “rationalist, I’m so much smarter than you, so I know better” asshole Scott Alexander hasn't been skewered yet for promotion of gambling (prediction markets) that people like Trumps son are making millions of on https://www.theguardian.com/us-news/ng-interactive/2026/may/...

We elected con man so we're becoming a swindler nation. In a just world the Andreesen David Sacks types would be in jail stripped of all their money.

Just out of curiosity, what is your legitimate criticism of Adreessen and Sacks?

Re: The AI Superforecasters Are Here

#29

Post by VMG »

Here’s my dystopian sci-fi scenario:

As prediction markets already show, forecasts can influence the outcomes they are trying to predict.

What happens when these models become extremely accurate and widely trusted? A forecast like “Will there be a war between countries A and B?” may itself affect whether the war happens.

If the model says there is a 1% chance of war, little changes. But if it says 90%, governments, markets, militaries, and the public may react: capital flees, troops mobilize, diplomatic trust collapses, and each side starts preparing for the other side’s preparation. The prediction helps make itself true.

The same feedback loop could apply to bank runs, market crashes, civil unrest, elections, and corporate failures.

At some point, the most accurate forecaster may become less like an observer and more like an actor with enormous power over the system it predicts.

Re: The AI Superforecasters Are Here

#30

Post by guardiangod »

Given a powerful enough AI with enough outlays (eg. a Playwright instant), can an AI cheat by influencing the real world and self-fulfil its prediction?

Eg. Is corp ABC's stock price going to go down by 10% in 30 days?

Orchestrate a disinformation campaign on the 29th day to tank ABC's stock price.

Post reply on HN