Interview
Richard Craib: WallStreetBets, Numerai, and the Future of Stock Trading | Lex Fridman Podcast #159
- Crowdsourced, AI-driven systems utilizing distributed networks are predicted to disrupt centralized financial power structures, potentially shifting money management from elite hedge funds to coordinated groups of individuals empowered by advanced machine learning tools.
- A primary risk involves the unpredictability of such decentralized freedom, where emergent behaviors may oscillate between constructive progress and chaotic outcomes driven by charismatic leaders or the potential for bad actors to manipulate anonymous groups.
- Specific to Wall Street Bets, historical analysis suggests coordinated short squeezes could have caused significant cascading deleveraging and damage to hedge funds; if such attacks had continued, the scenario posits a high probability of every hedge fund collapsing.
- While the destruction of hedge funds is viewed as potentially destabilizing due to their role in providing liquidity for venture capitalists, the continued growth of retail coordination is expected to lead to the selection of new targets for market attacks.
- Numerai outlines a strategic plan to manage global capital by aggregating data and talent, with a commitment to 10x the amount of data provided annually for at least the next 10 years to establish the world's most comprehensive financial dataset.
- Numerai's operational framework includes an ensemble model with a 30-day prediction horizon and a 3-4 month holding period, while incorporating unique 7-day signals derived from natural language sentiment scraped from sources like Wall Street Bets.
- The platform expects to automatically synthesize new machine learning techniques into its models and anticipates an increase in the number of users and machine learning researchers participating in its competitions.
- A core prediction asserts that a narrow intelligence will eventually win all money in the stock market, a milestone expected to occur sooner than the arrival of artificial general intelligence (AGI).
- Future expansion of Numerai includes importing signals from orthogonal data sources and integrating staking protocols, such as cryptographically securing comments to mitigate the negative behaviors associated with anonymity.
- Risks associated with fully AI-run markets include the emergence of bad actors where human morality might be silenced, contrasting with the potential for Numerai to become powerful without being dangerous if governed by its specific data and talent acquisition strategies.
- The underlying data strategy involves providing access to high-quality, obfuscated datasets spanning 15 to 20 years of history, valued at approximately one million dollars annually, which serves as the foundation for training models to avoid volatility from over-exposure to single features.
- Broader societal and individual outlooks suggest that comfort can diminish ambition, whereas realness, honesty, and boldness are expected to drive better outcomes, despite the inherent risks of making mistakes, with a belief that humans must avoid excessive comfort to achieve great things.