I think Watson was messing with the final question....
Spoiler!
Was about US Cities and a particular airport and Watson guessed "Toronto??????"....but he only bet $947 ahahahhahah
The 947 would come from his strategy algorithim where it came up with that the risk of answering incorrectly was greater than the benifit of wagering a lot.
I remember Ken attributed most of his success to his buzzing in ability. He claimed that most players know the answers but aren't as good as him at buzzing in (which I think is just modesty as he knows an inhuman amount of info).
He only means in comparison to other Jeopardy contestants.
There was a 60 minutes about memory freaks. Normies don't have a hope in hell of beating the super memory freaks. There were 6 or 8 of them and some of them said they make a living kicking ass on game shows.
Compared to your average person Ken is not only a genius but kicks ass at the buzzing in.
I think Watson was messing with the final question....
Spoiler!
Was about US Cities and a particular airport and Watson guessed "Toronto??????"....but he only bet $947 ahahahhahah
Spoiler!
If you look at Watson's 2nd and 3rd answers a lot of them make no sense. I suspect you'd get pretty weird answers most of the time if Watson didn't have the right answers so often.
Edit: Watson seems to struggle to identify absolute constraints... it doesn't properly eliminate answers that it should have 0% confidence in.
First, the category names on Jeopardy! are tricky. The answers often do not exactly fit the category. Watson, in his training phase, learned that categories only weakly suggest the kind of answer that is expected, and, therefore, the machine downgrades their significance. The way the language was parsed provided an advantage for the humans and a disadvantage for Watson, as well. “What US city” wasn’t in the question. If it had been, Watson would have given US cities much more weight as it searched for the answer. Adding to the confusion for Watson, there are cities named Toronto in the United States and the Toronto in Canada has an American League baseball team. It probably picked up those facts from the written material it has digested. Also, the machine didn’t find much evidence to connect either city’s airport to World War II. (Chicago was a very close second on Watson’s list of possible answers.) So this is just one of those situations that’s a snap for a reasonably knowledgeable human but a true brain teaser for the machine.
Location: In my office, at the Ministry of Awesome!
Exp:
The people that are arguing that this isn't really a computer beating people at Jeopardy because it can read the clues faster don't really get the signifigance of what Watson (or IBM with Watson) is doing. The fact that he's coming up with the right answer the vast majority of the time really is amazing.
__________________
THE SHANTZ WILL RISE AGAIN.
<-----Check the Badge bitches. You want some Awesome, you come to me!
The Following 3 Users Say Thank You to Bring_Back_Shantz For This Useful Post:
Trebek: And now onto Watson. I'm sure everyone is curious to put down what it answered.
...
Trebek: "What is Toronto?????" No, I'm sorry, that is not the correct question. And also note the number of question marks directly correlates to Watson's lack of confidence in its answer. And you wagered:
...
There is a lot of "guessing" about how Watson works.
He doesn't just buzz in immediately, he has to be sure of his answer before buzzing in. On average it was taking him 2.6 seconds. Not sure if that has been improved since his debut on the real Jeopady.
This is a great video on how he works and what an amazing machine he is. Playing Jeopardy is extremely complex. Watson isn't just an encyclopedia. That's not how Jeopardy works.
The people that are arguing that this isn't really a computer beating people at Jeopardy because it can read the clues faster don't really get the signifigance of what Watson (or IBM with Watson) is doing. The fact that he's coming up with the right answer the vast majority of the time really is amazing.
I find I'm kind of watching the question to see if it has anything that may be misleading to Watson.
So if I see a question that is fairly straight forward with a couple clues in it, I expect Watson to get it correct. It is the ones with a bit of a red herring that Watson still figures out that I find interesting.
__________________
"The problem with any ideology is that it gives the answer before you look at the evidence."
—Bill Clinton
"The greatest obstacle to discovery is not ignorance--it is the illusion of knowledge."
—Daniel J. Boorstin, historian, former Librarian of Congress
"But the Senator, while insisting he was not intoxicated, could not explain his nudity"
—WKRP in Cincinatti
Whats next for Watson if he beats the two most legendary players of all time? Move on to wheel of fortune?
__________________ "In brightest day, in blackest night / No evil shall escape my sight / Let those who worship evil's might / Beware my power, Green Lantern's light!"
Seems to me that we're playing a game that isn't that hard to win at.
I'm not sure what you're trying to say here, espeically by linking to that wikipedia article, but yes it is amazing.
What they've done is build a computer that can recognize natural language including slang, puns, and other things that even other humans have trouble with, if they are new to the language, and have it drawing conclusions from that. That's something that no one has been able to do before, and it's a pretty big step forward.
And yes, that game really is that hard to win at. Being able to piece together the right answer based on clues, not search parameters, is a pretty difficult thing to do.
If it's as easy as you claim it is, then why has no one done it before, and why did it take some of the best engineers at one of the largest computer companies in the world 3 years to accomplish it after they came up with the idea?
__________________
THE SHANTZ WILL RISE AGAIN.
<-----Check the Badge bitches. You want some Awesome, you come to me!
Last edited by Bring_Back_Shantz; 02-16-2011 at 01:30 PM.
The Following 2 Users Say Thank You to Bring_Back_Shantz For This Useful Post: