Post by Eric Thain
Venture Partner | LLB | Chair, AI Society | Digital Sovereignty Strategist
AI isn’t thinking. It’s guessing. 🧠⚠️ "AI is extraordinarily confident in its own abilities. And it's confident in telling you things that are potentially completely incorrect." In this clip from the latest Chair Sessions, Mario Camaj, AI governance expert in New York drops a truth bomb that every leader needs to hear: LLMs are trained to make next-token guesses. They don't have human systematic intelligence —they have an algorithmic confidence trick. Without strict guardrails and defined agents, letting AI run wild in your system means letting a hyper-confident guessing machine do whatever it wants. Watch the full breakdown on YouTube. Link in comments! 🎬 Artificial Intelligence Society of Hong Kong #AIGovernance
Video Content