Imagine a brilliant lawyer who’s memorized every law, statute, and precedent, but has never stepped into a courtroom. They’ve read every contract template, yet have no idea which clauses actually hold up in a messy, real-world dispute. This is the gap between theory and practice - or between explicit and tacit knowledge.
Tacit knowledge is this elusive expertise that comes from experience and is difficult to capture in writing.Today’s models are unparalleled at processing explicit knowledge (facts, data, etc.). However, they might be capped in scenarios that require more tacit knowledge.
The trillion-dollar question: How do we teach AI the unwritten knowledge? My guess is that it will require AI systems that learn not just from datasets and simulated environments, but learn through continuous interaction in real-world environments, feedback loops with experts, or even “apprenticeships” alongside humans.Imagine an AI that hasn't just read every legal textbook and documented legal case but has participated in millions of courtroom sessions and negotiations.
OpenAI and others are likely racing to crack this.
The trillion-dollar question: How do we teach AI the unwritten knowledge? My guess is that it will require AI systems that learn not just from datasets and simulated environments, but learn through continuous interaction in real-world environments, feedback loops with experts, or even “apprenticeships” alongside humans.Imagine an AI that hasn't just read every legal textbook and documented legal case but has participated in millions of courtroom sessions and negotiations.
OpenAI and others are likely racing to crack this.