Forecasting agents are the next frontier: RL on verifiable rewards, says ibab, worth $10k+/month
ibab · x · 2026-11-14
ibab argues that after coding agents' success, forecasting and decision-making are the new frontier. Forecasting is amenable to large-scale RL with verifiable rewards: agents are tested on historical data, predicting past markets and events with access only to prior events, and can be trained to superhuman level. CEOs and politicians will find such tools irresistible—willing to pay $10,000+/month—if agents are demonstrably smarter. He's building open-source forecasting RL recipes at riveraiinc so any organization can own its decision-making agents.
More from AGI Musings
- 'Messy Jobs' makes Bloomberg's best books of 2026, with praise from Orszag and Brynjolfsson — TaniaBabina · 2026-11-14
- Commentary: Regulating open source hands China the win in AI's open future — StewartalsopIII · 2026-11-14
- Stanford COLM 2026 paper: LLM-generated teaching content is far less diverse than human work — HariSubramonyam · 2026-11-14
- Many AI disagreements boil down to one question: is superintelligence real — binarybits · 2026-11-14
- Researchers banter: will human-generated pretraining data stay scarce? — DimitrisPapail · 2026-11-14
- Dev argues golden-age personal computing meets LLM agents for a new creative wave — jakedahn · 2026-11-14