Libratus, Carnegie Mellons poker-playing AI, is damn good. It easily routedfour of the worlds best poker players back in January over 120,000 hands of No-Limit Texas Hold-um. So it comes as no surprise that a team of venture capitalists, entrepreneurs and engineersplaying against Lengpudashi,Libratus bigger, badder brother, would face a similar fate. Plot twist: they lost, badly.
Now thats not meant to be unfair to Alan Du, the VC and poker pro leading Team Dragon against Lengpudashi in this weeksChinese exhibition match.The groups approach was actually quite valid.Rather than treatLengpudashi as a human and try to beatit with traditional strategy, Team Dragon went into the match with the aim of playing it like a machine. Unfortunately, it seems like a dash of game theory and an understanding of machine intelligence was of no real help duringthe match-up.
In a Bloomberg piece, Durelated poker to venture capital in an effort to express his teams preparedness. Yeteach successive day of the five-day match, Team Dragon sunkdeeper into a hole. This is because merely being knowledgeable about uncertainty isnt enough to take on finely tuned machines in an environment to which theyve already become accustomed.
Humans hold the advantage in startup investing because the ecosystem is too complex to model effectively. But if we found a way to do it,that system would be able to make macro bets that would seem nonsensical to humans making micro moves. Once machines are able to tilt the game in favor of their computational power, its pretty much game over for humans.
Theres a case to be made that the worlds foremost experts on machine learning and game theory might have stood a chance against the oldLibratus, butLengpudashi is even better.That said, Id stillpay good money to watch greats like Richard Sutton andJoseph Halpern take onLengpudashi with a heck of a lot of intentionality and a strong grasp of reinforcement learning and managing uncertainty.