Watson = the ultimate bot?
 

Interesting article in Slate




It may be surprising to learn that it's much easier to build a computer that can win at Jeopardy! than one that cleans up at the poker table in real-world situations. The quiz show, after all, can draw from any subject on God's green earth, while card games are based on 52 discrete units that interact with presumably calculable probabilities. Good Texas Hold 'Em players can estimate the odds of completing this or that hand based on the cards they see in their hands and on the table. So why can't a computer kick some ass at the casino?

I watched those jeopardy episodes and I was very impressed... The computer completely dominated the 2 greatest jeopardy champions ever. I thought a few of the computers wrong answers were funny and the daily double wagers were funny (2,247 lol). I was very impress with it.

Probably because IBM is behind solving jeopardy. Give it the same leeway with poker and it would destroy. As for watson, it holds a huge advantage with buzzing in which is probably the biggest aspect of jeopardy. Developers should have given it a random buzz in time according to jeopardy players of past.

There is a bot playing at bellagio at HU LHE and from all accounts it seems to do well. That is covered in the article though.

Not sure why it said bots should be used in the financial markets. I imagine they already, but I don't know anything about the subject other than minor knowledge of the subject.

The impression I got was how the human mind is still far superior.

See, as impressive as Watson was, it was simply brute force. 100 State of the art high speed servers rigged in sequence. A bludgeon of a search engine.

Example: The Final Jeopardy! clue in the first game:

category: US Cities
answer: Its largest airport is named for a World War II hero; its second largest, for a World War II battle

Easy clue for a human. We can discard irrelevancies here. Since a "2nd largest airport" is mentioned, we automatically KNOW we're looking at maybe the 5 largest cities. We identify the key to the clue as a WWII BATTLE and almost immediately come up with Chicago-Midway. It is irrelevant to our decision making whether or not we have ever heard of a war hero named "O'Hare". We, as humans, write down "What is Chicago" with complete confidence.

Watson, who KNOWS NOTHING, begins by parsing the names of airports in comparison to names of WWII medal winners, since the machine can't eliminate that part of the clue as meaningless since it is not intelligent. This initial search is a huge undertaking even for this mega Google on steroids, and it runs out of time, and probably because it has found a war hero named "Pearson", it answers "What is Toronto"

The humans here easily dominate the machine when the task actually involves intelligent decision-making.

SECOND POINT: Watson dominates because from the point of answer confidence to ring in, the humans have about a 0.3 second latency. The brain must signal the thumb to move and the thumb must push the button. Watson simply send a speed-of-light impulse to ring in. Even if the human gets to the answer 0.2 seconds before Watson, he can't ring in fast enough. watch the looks on Jennings' and Rutters' faces. They can't ring in fast enough. Build that same latency into Watson's ring-in function and the game is much closer. The humans may even win.

Watson was impressive and cool to watch, but is it in Jenning's words "Our new overlord", a CYbernetic Life fOrm Node?

Not even close.

yeah that first final was pretty funny since he did not name a U.S. City. Also like you said, its all about how fast u can ring in. 95% of the time on any given jeopardy question, all 3 players know the answer, its just a matter of who rings in first. You could tell that Jennings was getting pretty pissed about not ringing in fast enough lol.