The AI Superforecasters Are Here

A flashy new claim has entered the chat: AI “superforecasters” — basically souped-up chatbots trained to research deeply and make very specific predictions — are allegedly making serious money on betting-style prediction markets. In the article, one founder boasts an AI turned $35 into $2 million on Kalshi, while another says their system is beating the stock market. The demo itself sounds almost sci-fi: the bot spun up mini-helper bots, read a pile of websites, and concluded there was only a 7% chance the US could cut respiratory infections like colds in half by 2040.

But in the comments? Oh, the crowd was not ready to crown the robot oracle. The biggest mood was pure side-eye. One commenter basically said: if these systems are really that good, why are the founders selling subscriptions instead of quietly getting rich? Another immediately piled on with the brutal arithmetic: if the AI costs thousands a year to run, is there even any money left after expenses?

Then the thread took a hard turn from skepticism into full moral outrage. Critics slammed prediction markets as dressed-up gambling and accused the whole scene of being powered by greed, ego, and questionable people. One especially furious commenter went scorched-earth on the article’s tone, while another brought in Nassim Taleb’s classic anti-forecaster energy with the deadpan jab: “get another job.” That sparked its own mini-dunk war, with someone snapping back, “Tell me you don’t understand Taleb...” So yes: the bots may be forecasting the future, but the comments were busy forecasting a backlash.