>the humans repeatedly press the button >Watson still beats them because he can just spam the electrical impulse for buzzing in >people are amazed it won I don't get it. If the competition was "fair", ie limiting the rate anybody would buzz in, the humans would have wiped the floor with it.

File: 21 KB, 350x209, watson.jpg [View same] [iqdb] [saucenao] [google]

Anonymous Sat Feb 19 14:02:57 2011 No.2565741 [Reply] [Original]

>the humans repeatedly press the button
>Watson still beats them because he can just spam the electrical impulse for buzzing in
>people are amazed it won
I don't get it. If the competition was "fair", ie limiting the rate anybody would buzz in, the humans would have wiped the floor with it.

>>	EK !EKFQOBUFnQ Sat Feb 19 14:04:57 2011 No.2565751 what is the aim of he game? i have never seen this.

>>	Anonymous Sat Feb 19 14:04:57 2011 No.2565753 Butthurt human detected..bow down to your new masters inferior being!

>>	Anonymous Sat Feb 19 14:05:57 2011 No.2565758 >electrical impulse for buzzing in Actually, it was a servo that mechanically pushed the button, just like the human thumb. Thanks for playing, though.

>>	Maths Anonymous Sat Feb 19 14:05:57 2011 No.2565761 File: 99 KB, 1280x660, hlp.png [View same] [iqdb] [saucenao] [google] Hello Mathmatifags. CIRCLE THEREOMS. NOBODY FUCKING KNOWS. Prove me wrong with answers. pic related; questions

>>	Anonymous Sat Feb 19 14:07:57 2011 No.2565775 >>2565758 In both the practice round and the real round, you could see Ken and Rutter repeatedly pressing the button but the computer still came in first every single time. The only time it didn't buzz in was when it didn't know.

>>	Anonymous Sat Feb 19 14:08:57 2011 No.2565777 >>2565753 I, for one, welcome our new robotic overlords.

>>	Anonymous Sat Feb 19 14:09:57 2011 No.2565782 >>2565775 So you're saying it was too good? That was kind of the point... given the same rules as the other players it was obviously better at the game. It played better, harder, faster, stronger.

>>	Anonymous Sat Feb 19 14:09:57 2011 No.2565784 >>2565775 You know why that is? because the computer is smarter than the human, and can therefore think of the correct answer first

Anonymous Sat Feb 19 14:10:57 2011 No.2565788

>>2565775
>hurr durr
The buzzers are locked until the question is completed. Watson doesn't attempt to answer until the question is completed. The humans just spammed the button, and likely hit it a few milliseconds later than Watson.
Isn't simple logic amazing?

Anonymous Sat Feb 19 14:10:57 2011 No.2565789

The point of the game was to show that a computer could analyze a question and give an answer with the same speed as a human. If you think a supercomputer who has to search through and analyze 15TB worth of text is somehow cheating because it can press the clicker faster than a human, you're fucking retarded.

Anonymous Sat Feb 19 14:13:57 2011 No.2565802

>>2565782
I'm saying it wasn't exciting as some apparently think it is.

It used its one advantage (spamming the button faster than any human can) to offset its weakness (knowing stuff). On a level playing field, the humans would have wiped the floor with it.

It's the same as those AIs that play starcraft. Sure they can micromanage thousands of units and do amazing things, but they still get beat by humans who can outstrategize it.

Anonymous Sat Feb 19 14:14:57 2011 No.2565810

>>2565788
>>2565784
Part of the technology used to win the Jeopardy! contest was the electronic circuitry that received the "ready" signal, then examined whether Watson's confidence level was great enough to relay the signal to the buzzer solenoid, and then actuate the solenoid. Given the speed of this circuitry compared to the speed of human reaction times, Watson's reaction time was faster than the human contestants, except when the human successfully anticipated (instead of reacted to) the ready signal.

>>	Anonymous Sat Feb 19 14:17:57 2011 No.2565818 It played by the rules, and won. Anyone saying it was "unfair" is butthurt to the point of irrationality.

>>	Anonymous Sat Feb 19 14:17:57 2011 No.2565823 >>2565802 >On a level playing field, the humans would have wiped the floor with it. Nope.

>>	Anonymous Sat Feb 19 14:18:57 2011 No.2565828 >>2565823 >>2565802 Also, what the fuck is a "level playing field"? If you're going to swim again Michael Phelps, is it a "level playing field" is you tie his arms behind his back first? Is that the reasoning you're using?

>>	Anonymous Sat Feb 19 14:19:57 2011 No.2565835 >>2565828 >If you're going to swim against Michael Phelps, is it a "level playing field" if you tie his arms behind his back first? Is that the reasoning you're using? fix'd

>>	Anonymous Sat Feb 19 14:20:57 2011 No.2565837 >>2565761 I wouldn't mind the answers to these actually. I completely forgot all of this shit apparently.

>>	Anonymous Sat Feb 19 14:22:24 2011 No.2565847 >>2565823 how would you level the playing field exactly?

>>	Anonymous Sat Feb 19 14:24:24 2011 No.2565857 i could beat watson its only a computer. you cant explain that

>>	Anonymous Sat Feb 19 14:24:57 2011 No.2565856 So the solution would be to artificially limit somebodies capacity to buzz in?

>>	Anonymous Sat Feb 19 14:25:24 2011 No.2565864 File: 43 KB, 300x300, billo'reilly.jpg [View same] [iqdb] [saucenao] [google] >>2565857 why do i keep forgetting my pictures? you cant explain that

>>	Anonymous Sat Feb 19 14:26:24 2011 No.2565868 >>2565856 yes. like people who cant quick draw or putting Husein bolt in clown shoes to level the playing field. he should give others a chance

>>	Anonymous Sat Feb 19 14:26:57 2011 No.2565869 >>2565864 You forget your pictures. Watson never forgets. You can't explain that.

>>	Anonymous Sat Feb 19 14:27:34 2011 No.2565870 watson was bullshit paid advertising for IBM did anybody think that a computer couldn't beat humans to a fucking electronic buzzer?

Anonymous Sat Feb 19 14:28:57 2011 No.2565873

>>2565847
I don't think that has any meaning anyway. Jeopardy has a set of rules. They were all followed. Watson won.

If it means finding a ruleset that is harder for computers without being harder for humans, I think that's worth looking in to as a next step. But all this "baaaaaw unfair" bullshit is really tiresome.

>>	Anonymous Sat Feb 19 14:29:57 2011 No.2565877 >>2565868 So, ideally, we should measure how quick every contestant buzzes before the show proper, and then apply a handicap to those who are faster than average?

Anonymous Sat Feb 19 14:29:57 2011 No.2565880

>>2565741
its not even the fact the robot can spam the button faster than the humans but can give the correct answer after searching his huge database in less than a few seconds and than properly say that answer with the correct english nuances is what was trying to be fulfilled.
Go on go google an open ended question from Jeopardy and tell me if you get 1 answer and not 7,200,000,000 in less than .0003 seconds.
what IBM did was amazing and you're just not going to accept that a robot beat a human at a very intense trivia game.

>>	Anonymous Sat Feb 19 14:30:24 2011 No.2565882 >>2565870 i thought it had a piston hand

>>	Anonymous Sat Feb 19 14:30:57 2011 No.2565881 >>2565868 It's not fair being being able to press a button quickly is not the point of the game. Running quickly is the point of the 100/200m, so Usain Bolt in clown shoes isn't a valid comparison.

>>	Anonymous Sat Feb 19 14:30:57 2011 No.2565885 >>2565877 How would you be able to objectively say someone is slower than somebody else, rather than intentionally skewing his answers?

>>	Anonymous Sat Feb 19 14:31:34 2011 No.2565889 >>2565873 wrong. wrong. wrong. Watson DID NOT follow the same rules as for humans. It DID NOT press a buzzer. It had a severe electronic advantage. It was paid advertising for IBM...deal with it

>>	Anonymous Sat Feb 19 14:31:57 2011 No.2565891 >>2565881 >It's not fair being being able to press a button quickly is not the point of the game. Full retard. It IS a VERY important part of the game.

>>	Anonymous Sat Feb 19 14:32:24 2011 No.2565896 >>2565877 yeah basically if youre fast you have to run to buzzer or somthing and the slow ones get to have the puzzer in thier hand so its FAIR or somthing

Anonymous Sat Feb 19 14:32:57 2011 No.2565899

>>2565889
>Watson DID NOT follow the same rules as for humans. It DID NOT press a buzzer.
You're wrong, it had to press a mechanical button just like the other players.

You just pulled that out of your ass, didn't you? Have you no intellectual integrity?

>>	Anonymous Sat Feb 19 14:33:57 2011 No.2565902 goddamn, the amazing part isn't that it can win at jeopardy. The amazing part is that it can ANSWER QUESTIONS all by itself.

Anonymous Sat Feb 19 14:34:57 2011 No.2565905

>>2565885

Exactly. The system we have now is perfectly fair.

The contestant has to be the first with the right answer. Being first is part of the competition, and being accurate is part of the competition.

The only reason this is contentious is that it is something that machines have only recently gotten better at. We don't let cars compete in the Olympics of the Kentucky Derby, right? But when cars were just invented, there would have been some interest in seeing whether they could outrun a person or a horse. This is that.

Anonymous Sat Feb 19 14:36:24 2011 No.2565923

>>2565902
exactly

we already know computers can beat 99% of people in anything speedwise.

i cant even win against ryu in hard mode at street fighter its just too quick.

so fucking what its fast blah blah

but it can fucking give me an answer to any question in a split second and it be right

fucking butthrt fags ITT bawwing over buzzer speed

they need to read this

http://www.kurzweilai.net/the-buzzer-factor-did-watson-have-an-unfair-advantage

the stfu and gtfo

Anonymous Sat Feb 19 14:36:57 2011 No.2565922

It's (relatively) easy for a machine to press a button faster than a human.

It's (usually) much easier for a human to process a query in natural language and respond in kind.

So if it's unfair that Watson could press a button faster, it's also unfair that the humans can comprehend natural language so effortlessly.

The fact that Watson is so good at processing natural language (within the domain of Jeopardy questions) that he can beat human opponents was kind of the whole point of the IBM challenge -- they were showcasing how far AIs have advanced in natural language processing.

>>	Anonymous Sat Feb 19 14:38:57 2011 No.2565928 >>2565905 This. "Fair" is nothing more or less than applying the same rules to everyone. If you think a given contest is uninteresting, then change the ruleset. But applying artificial handicaps to come players and not others is unfair.

>>	Anonymous Sat Feb 19 14:38:57 2011 No.2565930 >herp derp computerr dint win fair n square It was meant to show off the natural language processing power of Watson. The Jeopardy game was just a big advertisement.

Anonymous Sat Feb 19 14:39:24 2011 No.2565933

>>2565905
well the whole automated train tracks machine is the best example of that

i cant remember but some slave beat the machine but died the posint was made the machine kept the pace and laid as much track as him and didnt die

http://en.wikipedia.org/wiki/John_Henry_(folklore)

>>	Anonymous Sat Feb 19 14:39:57 2011 No.2565931 I would have preferred Watson having to actually read the clue and process the optical input.

Anonymous Sat Feb 19 14:39:57 2011 No.2565935

>>2565922
Exactly. Where are all the people bitching about the millions of years of development time that humanity has had to produce the human brain, where Watson only had a few years of design time?

Fuck artificial restrictions. The rules of jeopardy apply equally to all players. If you don't like the game, change the rules, but apply those rules to all the players equally.

Anonymous Sat Feb 19 14:42:24 2011 No.2565948

>>2565928
games are the handicap

the whole point of a game is to get three people who have exactly the same thing and the only difference is thier ability to work that thing better than the other competitors

boo hoo a computer won because it was better at the game

get over it already.

i mean fuck you dont get mad when someone is faster than you at a game like streetfighter

ok well i do but just because i feel i should be beeter and faster than i am.

>>	Anonymous Sat Feb 19 14:44:34 2011 No.2565955 did watson have to listen to the clues or read the clues? if it had electronic information transmitted to it, then it wasn't playing by the same rules as humans.

>>	Anonymous Sat Feb 19 14:44:57 2011 No.2565954 Here's a full practice round of Watson for PBS. Watch him fucking get rhyming. Now tell me it's not fucking amazing for a computer to figure this shit out. http://www.youtube.com/watch?v=YgYSv2KSyWg http://www.youtube.com/watch?v=dr7IxQeXr7g

>>	Anonymous Sat Feb 19 14:45:24 2011 No.2565963 >>2565955 ah now this is a good point. i dont know.

Anonymous Sat Feb 19 14:45:57 2011 No.2565964

It amazes me how many people miss the point of Watson and jump straight to the buzzer issue. The fact that a computer can understand a great deal of the intricacies of human language is a feat in and of itself. If you're upset because of the "buzzer effect", you completely missed the point.

Remember when a computer with the processing power of today's smartphones had to be housed in a whole room? Imagine 30-50 years from now, when supercomputers like Watson can fit in a desktop tower. You'll be able to walk into your home and control various settings in your house by talking to a 3D hologram. Instead of getting caught up in trivial shit about buzzers, why not think about how FUCKING AWESOME watson is?

related: http://www.youtube.com/watch?v=H7K0GQPnmDI

>>	Anonymous Sat Feb 19 14:48:34 2011 No.2565981 >>2565964 what upsets people is giving the computer a different set of rules. read this>>2565955 if it did have different rules, it was NOT playing JEOPARDY!, it was advertising for IBM

>>	Anonymous Sat Feb 19 14:48:57 2011 No.2565976 Also, this guy has the first two days if you haven't seen the actual epsides yet. http://www.youtube.com/watch?v=ZLdkJpAtt1I

>>	Anonymous Sat Feb 19 14:50:57 2011 No.2565983 >>2565955 Clues were sent as a text file. Now, that'd be somewhat fair if, say, it was sent the same instant the buzzers were activated.

Anonymous Sat Feb 19 14:52:57 2011 No.2565995
File: 24 KB, 340x426, 1222306622874.jpg [View same] [iqdb] [saucenao] [google]

When Deep Blue won at chess, no one cried that he didn't have a camera pointed at the chess board and interpret the piece locations. No one cried that his opponent's moves were fed to him.

Because everyone fucking got that the point of the experiment was to see if Deep Blue could even fucking compete 1) against a human at all 2) at a decent speed.

Same with Watson. Next robot will look like T-101 and have cameras for eyes.

>>	Anonymous Sat Feb 19 14:53:34 2011 No.2566002 >>2565983 Watson WAS NOT PLAYING JEOPARDY! It was advertising for IBM....deal with it problem solved, thread over anything below this line is trolling: --------------------------------------------------

>>	Anonymous Sat Feb 19 14:54:34 2011 No.2566005 >>2565995 BULLSHIT chess is not a timed game, at least in the matches you refer to.

>>	Anonymous Sat Feb 19 14:55:57 2011 No.2566009 The questions were available to Watson at the same time they became available to the meatbags.

Anonymous Sat Feb 19 14:55:57 2011 No.2566010

Watson had significantly more correct answers than its two opponents, you can find the ratio to questions answered correctly/money won online.

Watson also had very abstract betting, would give answers to questions so far off a human wouldn't consider (IE montreal isn't a US city), and even repeated wrong answers. These are obvious, measurable disadvantages. (The IBM engineers should have known to add that Watson couldn't give repetitive wrong answers)

>>	Anonymous Sat Feb 19 14:56:34 2011 No.2566011 >>2565995 BULLSHIT PART 2 it wasn't an experiment, like the deep blue, it was fucking advertising get over it comp sci majors, you were whored out for advertising

>>	Anonymous Sat Feb 19 14:56:57 2011 No.2566014 So it was advertising for IBM. WHO CARES? They did something awesome and deserve recognition for it. All the money went to charity anyway.

>>	Anonymous Sat Feb 19 14:57:57 2011 No.2566020 >>2566005 K vs. Deep Blue was timed (though K forfeited a few matches before checkmate)

>>	Anonymous Sat Feb 19 14:58:34 2011 No.2566026 >>2566009 the humans didn't have the questions piped directly into their brain, you're wrong >>2565983 Watson WAS NOT PLAYING JEOPARDY! It was advertising for IBM....deal with it

>>	Anonymous Sat Feb 19 14:59:34 2011 No.2566028 >>2566020 your comparing apples and oranges

Anonymous Sat Feb 19 14:59:57 2011 No.2566029

>>2566005
>chess is not a timed game, at least in the matches you refer to.
Yes they were.

http://en.wikipedia.org/wiki/Deep_Blue_versus_Garry_Kasparov
>The first game of the 1996 match was the first game to be won by a chess-playing computer against a reigning world champion under normal chess tournament conditions, and in particular, normal time controls.

>>	Anonymous Sat Feb 19 15:00:34 2011 No.2566036 >>2566014 oh i don't know maybe google? maybe all the people at home who were duped? they didn't do shit that was special except buy advertising time on JEOPARDY!. they deserve ZERO recognition

絶望先生 Sat Feb 19 15:01:57 2011 No.2566040
File: 37 KB, 500x281, 69.jpg [View same] [iqdb] [saucenao] [google]

>>2565954
Watson did not "get" rhyming, it broke down the question by finding patterns and googled the answer in its database. Watson is not an AI, it is a number crunching autonomous search engine with a large database of trivia and patterns found through computer learning. It is not intelligent in any sort of way, calling watson an AI is insulting to the people who are actually developing AI.

Anonymous Sat Feb 19 15:02:57 2011 No.2566044

>>2566010
>(The IBM engineers should have known to add that Watson couldn't give repetitive wrong answers)

There's no ractical way to do this, and even if they did that would be missing the point. The point wasn't to build a god-tier jeopardy playing machine.

Anonymous Sat Feb 19 15:03:57 2011 No.2566048

>>2565741

1. It is supposed to be handicapped to the level of the humans.
2. The rate with which is searches information is amazing none the less. However if you not a programmer or CS graduate you are not likely to understand cause hey it's magic right?

>>	Anonymous Sat Feb 19 15:03:57 2011 No.2566049 People who are still complaining about the timing advantage are missing the point. Read this: >>2565964 >>2565922 >>2565880

>>	Anonymous Sat Feb 19 15:04:34 2011 No.2566055 >>2566044 except it didn't actually play JEOPARDY! >>2565983 Watson WAS NOT PLAYING JEOPARDY! It was advertising for IBM....deal with it

Anonymous Sat Feb 19 15:04:57 2011 No.2566054
File: 44 KB, 500x307, 1220510012386.jpg [View same] [iqdb] [saucenao] [google]

It wasn't a game, dang nabbit! That whizzy computer thing cheated, he had no eyes! No ears!

IBM fed him the right answer. He HAD to cheat, computers don't work like that. I type into aol.com what I want to know, and the internet finds it for me. Why didn't the Internet get to play on Jeopardy?!!?!

BECAUSE THE INTERNET DIDNT PAY FOR ADVERTISING LIKE IBM!!!!!

HUMANS STILL #1!

>>	Anonymous Sat Feb 19 15:06:57 2011 No.2566064 >>2566055 It was both. Watson playing Jeopardy and the show being an IBM ad aren't mutually exclusive.

>>	Anonymous Sat Feb 19 15:08:34 2011 No.2566070 >>2566064 >>2565983 Watson WAS NOT PLAYING JEOPARDY! It was advertising for IBM....deal with it Listen nigger, when it reads the question and/or listens to the question, then it will be playing JEOPARDY! It did neither, It wasn't playing JEOPARDY!

>>	sage sage Sat Feb 19 15:11:57 2011 No.2566077 >>2566070 As was pointed out before, by that logic you would have to deny that Deep Blue played chess. Sage because you're immune to reason.

>>	Anonymous Sat Feb 19 15:12:57 2011 No.2566082 Retardation: The thread.

Anonymous Sat Feb 19 15:13:57 2011 No.2566084

Jeopardy DOES limit buzzing in several ways. For instance, if you buzz before the end of the question, your subsequent buzzes are delayed something like 3/10 of a second. I'm pretty sure that repeated buzzes don't do shit, either, which is why you sometimes see people banging hard on the trigger and never getting first.

Anonymous Sat Feb 19 15:14:34 2011 No.2566095

>>2566077
wrong. you are comparing apples and oranges.

deep blue was a real experiment and everyone involved knew all the parameters

Watson was a marketing scheme and the entire home audience was deceived into believing Watson was playing by the same rules as the humans. It wasn't. It was deception. It was marketing. It was not a special achievement of any sort.

>>	Anonymous Sat Feb 19 15:16:57 2011 No.2566103 >>2566095 Sorry, if you don't move the pieces you aren't playing chess. Where's the apples and oranges?

>>	Anonymous Sat Feb 19 15:18:57 2011 No.2566106 >>2566095 What deception are you talking about? Trabek spelled out all the parameters at the beginning of the show.

>>	Anonymous Sat Feb 19 15:20:34 2011 No.2566111 >>2566106 and Orson Welles explained the parameters very clearly before War of the Worlds.....nearly collapsed the fucking country. It was a huge deception. Watson was not playing JEOPARDY!

>>	Anonymous Sat Feb 19 15:20:57 2011 No.2566112 >>2566095 By the same token, Deep Blue was also a marketing scheme (which, obviously, it actually was). But that doesn't mean that it wasn't also an impressive technological achievement.

Anonymous Sat Feb 19 15:23:57 2011 No.2566119
File: 28 KB, 300x293, abe simpson.gif [View same] [iqdb] [saucenao] [google]

He wasn't playing! You want to know the rules? I'll tell you the rules!

Trebek gives each contestant a softball question about their life after the first round. Watson was not playing JEOPARDY!

>>	Anonymous Sat Feb 19 15:23:57 2011 No.2566122 >>2566111 This is a very stupid analogy. Welles pointed out what was to follow was a work of fiction. Trabek explained the ground rules of a game. How are these in any way comparable, other than men saying words over a broadcast medium?

>>	Anonymous Sat Feb 19 15:24:34 2011 No.2566125 >>2566112 Bullshit. Deep Blue was a real experiment and a real challenge to the greatest living human in that sport. Watson was deception and paid marketing 100%.

>>	Anonymous Sat Feb 19 15:25:34 2011 No.2566127 >>2566122 a brief disclaimer DOES NOT mean you aren't deceiving people.

>>	Anonymous Sat Feb 19 15:26:34 2011 No.2566133 >>2565983 Watson WAS NOT PLAYING JEOPARDY! It was advertising for IBM, It was deception of the home audience, it was comp sci being whored to the public....deal with it

Anonymous Sat Feb 19 15:26:57 2011 No.2566134

The part that amazes me is that humans could actually compete against it.
Think about all the effort and capacity needed to compete against 15TB of pure memory with no distraction or useless information like walking, breathing, seeing etc.
What's more, humans developed such an awesome machine, so FUCK YEAH HUMANITY!!!
(cant avoid fantasizing about the time we can implant an equivalent of Watson in our bodies and actually use it directly as part of our memory and cognitive processes)

>>	Anonymous Sat Feb 19 15:27:57 2011 No.2566137 File: 18 KB, 320x355, zap crying.jpg [View same] [iqdb] [saucenao] [google] why have you destroyed our beautiful thread?

Anonymous Sat Feb 19 15:28:57 2011 No.2566138

What they should have done was either
a) have every question be like Final Jeopardy, where everyone gets to answer; or
b) have a human pushing the buzzer for Watson, perhaps in response to a signal from Watson saying that it wanted to answer the question.
This would have eliminated the purely mechanical advantage that Watson had and made for a more interesting game. The point wasn't to show that a machine can react to a stimulus faster than a person can. It also wasn't to show that we can do image to text processing (which has been possible for many years). It was to compare the language processing abilities of an advanced machine (well, 2880 machines) to those of people. In this sense, the traditional Jeopardy setup was a failure, and it detracts from the accomplishments of the IBM team.

Anonymous Sat Feb 19 15:29:57 2011 No.2566143

>>2566127
If you listen to the broadcast, he's very explicitly states it is fiction and for entertainment. You don't have to say it's fiction every twenty minutes to remove your liability of gullible people.

Also, really /sci/? IBM invested millions upon millions into Watson, and the million$ prize went to charity. The people who watch Jeopardy are likely educated enough to make the decision to research a company before investing blindly into their stocks or products.

>>	Anonymous Sat Feb 19 15:29:57 2011 No.2566146 If you take letters of HAL and replace each letter with next one in alphabet you'll get IBM.

>>	Anonymous Sat Feb 19 15:30:57 2011 No.2566150 my computer can beat up your computer.

>>	Anonymous Sat Feb 19 15:30:57 2011 No.2566152 >>2566143 don't blame one sad doofus camping in this thread on the collective community of /sci/

>>	Anonymous Sat Feb 19 15:32:34 2011 No.2566157 >>2566138 Wrong the point was to advertise for IBM and most effectively deceive the american public. those changes you suggest would have compromised the actual goals.

>>	Anonymous Sat Feb 19 15:34:57 2011 No.2566163 Watson doesn't buzz until he knows the answer. That in itself OS interesting.

>>	Anonymous Sat Feb 19 15:35:57 2011 No.2566173 gonna be fucking awesome when i have one of these in my house in 20-30 years

>>	Anonymous Sat Feb 19 15:39:57 2011 No.2566202 IBM has fucking GREAT english text-to-speech code. Can we please get this shit in our smartphones?

Anonymous Sat Feb 19 15:40:57 2011 No.2566210

>>2566138
Maybe a different format would have showcased Watson's NLP abilities better, but I wouldn't go as far as to say it was a failure.

After all, if Watson sucked at NLP, his speed would not have been an advantage (and might have even been a disadvantage).

Also, don't forget that one of the advantages of a good NLP AI system is both NLP AND speed. Which Watson did a good job of showing off. (Some of the posters in this thread seem to think it did TOO good a job.)

>>	Anonymous Sat Feb 19 15:42:34 2011 No.2566216 >>2566163 >>2566173 >>2566202 ITT: viral marketers

>>	Anonymous Sat Feb 19 15:42:57 2011 No.2566215 File: 1.18 MB, 1023x578, 1298141520142.png [View same] [iqdb] [saucenao] [google]

>>	Anonymous Sat Feb 19 15:44:57 2011 No.2566220 >>2566216 butthurt samefag u mad?? but y?? u europeon? mad u insignificant? y u hate on i.b.m?

>>	Anonymous Sat Feb 19 15:46:57 2011 No.2566229 IBM was basically laying the groundwork for a service that it will offer for $$$ (or free + display ads). "Ask Watson", to compete with Google/Wikipedia.

>>	Anonymous Sat Feb 19 15:47:57 2011 No.2566232 >>2566229 "What is the meaning of life, the universe, and everything?"

>>	Anonymous Sat Feb 19 15:47:57 2011 No.2566233 >>2566229 lol if you think this that's akin to saying they developed big blue as a chess consultant

Anonymous Sat Feb 19 15:52:57 2011 No.2566257

>hupdurpdurp
Waston did run entirely on speech recognition software
He couldn't even read the questions, only hear
It's amazing how butthurt some people get about this.

Don't you see silly mortals, your time is over. How can you challenge a perfect, immortal machine?

Anonymous Sat Feb 19 16:00:57 2011 No.2566303

>Watson was too fast!
That's the whole fucking point you retards. The entire point of Watson was that IBM wanted to show it had created an AI that could crunch through natural language data fast enough to not only answer as fast as humans, but answer as fast, if not faster, as humans who are champion trivia show contestants.

>>	Anonymous Sat Feb 19 16:04:57 2011 No.2566316 File: 13 KB, 240x249, 1186048195074.jpg [View same] [iqdb] [saucenao] [google] >>2565954 >Your face as he got the rhyme thing

Anonymous Sat Feb 19 16:06:57 2011 No.2566322

>>2566210
Yeah, but it wasn't a comparison of the NLP speed. Undoubtedly the human contestants also knew the answers before they were allowed to buzz in... but their relatively shitty human reflexes meant they couldn't press the button fast enough to answer most questions. Maybe if they'd allowed them to ring in at any time before Trebek was done reading?

>>2566303
That's not what it showed, though. It mostly showed that the computer could click the button closer to the end of the reading of the question than could the humans. Perhaps the people *were* faster at language processing, but because of the format of Jeopardy, we'll never know.

>>	Anonymous Sat Feb 19 16:10:57 2011 No.2566338 EVERYONE IN THIS THREAD Watch this video now. http://www.youtube.com/watch?v=3G2H3DZ8rNc Then you can resume discussion.

Anonymous Sat Feb 19 16:14:57 2011 No.2566359

>>2566322
Yes, but the computer couldn't have clicked the button until it had come up with an answer. Thus, even if you assume it was only edging out the contestants because it could hit the button before them, that would still mean that it had arrived at the answer at essentially the same time as they did.

>>	Anonymous Sat Feb 19 16:16:34 2011 No.2566369 >>2566359 the computer had a longer processing time, and less actual processing to do. it wasn't playing JEOPARDY!. it was an advertising scam

>>	Anonymous Sat Feb 19 16:18:57 2011 No.2566375 >>2566257 >Waston did run entirely on speech recognition software No, he didn't. This is what I thought at first, which would have made it actually impressive.

Anonymous Sat Feb 19 16:18:57 2011 No.2566376

http://www.youtube.com/watch?v=dr7IxQeXr7g
Why does he always say "Nicely done, Watson" it's not like the machine can understand it
Maybe it's some trigger, so it understands it got the question right
Maybe he believes that if the machines ever rises up, he will be spared because he was always so nice to them

>>	Anonymous Sat Feb 19 16:22:34 2011 No.2566388 >>2566375 Clear proof that the american public was scammed by the Watson advertising HOAX. prepare for the viral marketers shitstorm responses

>>	Anonymous Sat Feb 19 16:22:57 2011 No.2566389 >>2566375 so, how does he get the questions then?

>>	Anonymous Sat Feb 19 16:23:34 2011 No.2566392 >>2566389 text file Watson was a marketing HOAX

>>	Anonymous Sat Feb 19 16:23:57 2011 No.2566393 >>2566338 I like this video, it shows some of his second choices. Especially say thank you in russian. Second choice? 'Thank you very much' With a confidence level above the threshold...

>>	Anonymous Sat Feb 19 16:23:57 2011 No.2566394 >>2566388 Jesus, does your butthurt ever stop? The machine won, humans lost, get over it

>>	Anonymous Sat Feb 19 16:24:57 2011 No.2566398 >>2566392 Could i get the answer how Watson got the questions from someone who isn't a huge toll? Preferably with sources

Anonymous Sat Feb 19 16:25:57 2011 No.2566401

>>2566322
>Yeah, but it wasn't a comparison of the NLP speed.

That's a fair point. Of course, that can also be said about competition between human players as well. (So, say, Ken might consistently formulate the correct answers faster than Brad, but Brad has a faster trigger finger and thus outscores Ken.)

Still, I wouldn't go as far as to call the format a failure, since it still shows how good Watson is at NLP -- perhaps not with as direct a comparison against humans as could have been done, but it was still able to establish that Watson is at least roughly at the same level (for correctly answering Jeopardy questions, at least).

So Watson has that, plus superhuman "reflexes". I'd say that's pretty impressive (and, potentially, marketable as an application in other domains).

>>	Anonymous Sat Feb 19 16:25:57 2011 No.2566403 Everybody knew watson got the question as a .txt when alex began reading. The test would have been fair if they abandoned the Jeopardy! rules and let anybody ring in as early as they wanted

>>	Anonymous Sat Feb 19 16:26:57 2011 No.2566408 >>2566398 text probably was mentioned in >>2566338

Anonymous Sat Feb 19 16:27:57 2011 No.2566411

I don't understand why everyone thinks the button thing is so fair. The rules on the show say that you can't buzz in until he's done reading the clue. Watson might have needed to push the button mechanically, but he probably has faster reaction times than any of the people, so he could more or less buzz in first every time unless the people happened to get super lucky with their timing. Regardless of whether or not the computer knew the answer before the people, it could safely assume that it WOULD know the answer relatively quickly (it is pretty smart) and buzz it in without being sure yet.

I'm definitely impressed by it, but saying the people were on a level playing field is a bit skewed. There were times where the people were mashing the button the instant Trebek finished reading the clue, so it was pretty obvious they figured out the answer as he was reading and got it "instantly." From there it would just be a question of who presses the button first, and I find the fact that the computer won 99% of the time a little uneven.

>>	Anonymous Sat Feb 19 16:27:57 2011 No.2566414 So how long until Google Answer where we can ask it questions in plain text and it will try to answer?

Anonymous Sat Feb 19 16:32:57 2011 No.2566425

>>2566411
See >>2565922

Also, it's fair in the sense that the rules were followed. It's not fair in the sense that Watson's abilities allowed it to dominate. It's like if the Miami Heat played a basketball game against a bunch of retarded 10 year olds. It's fair in the sense that both follow the same rules. It's not fair in that one side is clearly outmatched.

Anonymous Sat Feb 19 16:37:57 2011 No.2566440

>>2566359
Within a couple of seconds, but that could be as much as twice as long. Impressive, but not as impressive as the results would lead you to believe.

>>2566401
Oh yeah, I like the technology, no doubt. It's just that the inequity of the competition ruined it a little for me... I wanted to see the *limit* of its abilities. Sure, it can roughly compete with people, within an order of magnitude... but what's the best it can do under the same limitations as those people? Does it really get the answers faster, or is it just "fast enough"?

>>	STAR !!pdO9uKe1f+M Sat Feb 19 16:38:57 2011 No.2566442 Hey guys guess what! You can calm down, because machines aren't yet capable of creativity nor can they think for themselves. Jeez.

>>	Anonymous Sat Feb 19 16:39:57 2011 No.2566447 >>2565761 44, 25, 110, 37, 40, 102 I also hate circle theorems.

>>	Anonymous Sat Feb 19 17:43:57 2011 No.2566779 >>2566411 it's a very common jeporday strategy to ring before you have the answer, and watson couldnt do that

Anonymous Sat Feb 19 18:09:57 2011 No.2566948

>>2566401

Oh FFS,

Watson was given the .txt file in natural language after the buzzers were unlocked. That is to say, AFTER Trebek was done reading the question. The humans had considedrablly more time to read the question. Also of note is the the .txt files weren't just filling out some set set of XML style variables. They were just search queries for him. They were printed in the same natural language as the question everyone else got, and he had to READ and UNDERSTAND what it meant.

I can't believe people are butthurt that they didn't make technological breakthroughs in auido recognition AND AI Nat language parsing at the same time.

Anonymous Sat Feb 19 18:23:57 2011 No.2567046

>>2566948
>Watson was given the .txt file in natural language after the buzzers were unlocked. That is to say, AFTER Trebek was done reading the question. The humans had considedrablly more time to read the question.
I don't believe this is correct. IIRC he was given the text at the same time the questions is visually revealed, and Trebek begins reading. If he only got the question when the buzzers unlocked, how could he ever beat the humans to the buzzer?

But Watson is really impressive. Ignore the trolls, gentlemen.

>>	Anonymous Sat Feb 19 18:25:57 2011 No.2567058 troll threads everywhere!

>>	Anonymous Sun Feb 20 01:00:57 2011 No.2569824 bump

>>	Anonymous Sun Feb 20 01:19:57 2011 No.2569948 britfag here im watching the video, can someone tell me why they keep phrasing the answers as questions ?

>>	We Are All Anonymous Sun Feb 20 01:33:57 2011 No.2570030 for a computer to even remotely understand a question is amazing. sure computers can recognize words. but this is recognizing phrases. therefor WATSON = uber pwnage and human's = dumb niggers

Anonymous Sun Feb 20 02:26:57 2011 No.2570353

i would have liked to see it answer a visual clue or even a video clue, that would be much more impressive

cmon we already have voice reccognition and it said it could analyze thousands of photos to understand what you mean

mfw it thinks toronto is a u.s. city

>>	Inurdaes !V1sPhobos. Sun Feb 20 02:27:57 2011 No.2570366 http://arstechnica.com/media/news/2011/02/creators-watson-has-no-speed-advantage-as-it-crushes-human s-in-jeopardy.ars

>>	Anonymous Sun Feb 20 02:56:37 2011 No.2570511 Anybody notice that it was apparently all white people who worked on Watson? There was that one asian lady, and she even spoke perfect English. There's no way that team isn't like 80 percent Asian or Indian.

>>	Anonymous Sun Feb 20 02:58:34 2011 No.2570518 Watson received questions in a text file this is not JEOPARDY! when it can read or listen to the questions then it will be playing JEOPARDY! Watson was a MARKETING HOAX

Anonymous Sun Feb 20 02:59:57 2011 No.2570524

>>2569948
That's the quirk about Jeopardy. They give you the answer, and you have to give the question! It's crazy!

All that means is that it makes everything formatted really weird, like the question about some gymnast with one leg:

A: Fred Smith was famous because he won a gold medal or something, even though he had this handicap.

Q: What is he only had one leg?

and Watson got it wrong because he said "What is leg?"

>>	Anonymous Sun Feb 20 03:01:57 2011 No.2570530 >>2570353 voice recognition sucks and despite technical advances is plateauing. 80% it was, I think. yeah..wiki says it too.

>>	Anonymous Sun Feb 20 03:05:34 2011 No.2570542 Watson received answers in a text file this is not JEOPARDY! when it can read or listen to the questions then it will be playing JEOPARDY! Watson was a MARKETING HOAX

>>	Anonymous Sun Feb 20 03:05:57 2011 No.2570546 >>2570518 It's a non-issue. They could us OCR if they wanted with an easy 100% accuracy rate. It's just a pointless exercise, as they might as well just send a text file.

>>	Anonymous Sun Feb 20 03:09:34 2011 No.2570561 >>2570546 bullshit, that would take extra processing power and extra time and slow down Watson's buzzing rate

>>	Anonymous Sun Feb 20 03:10:34 2011 No.2570570 >>2570561 thus unfair advantage, thus marketing hoax NOT A JEOPARDY! GAME

>>	Anonymous Sun Feb 20 03:13:34 2011 No.2570586 ITT: IBM viral marketer's pwned by Watson's obvious hoax/cheating at JEOPARDY!

>>	Anonymous Sun Feb 20 03:13:57 2011 No.2570588 >>2570561 >>2570570 Bullshit. It would take 1 microsecond to OCR the text. It would make no difference at all.

>>	Anonymous Sun Feb 20 03:15:57 2011 No.2570600 File: 10 KB, 263x263, 1296428689503.jpg [View same] [iqdb] [saucenao] [google] >>2570511 Teehee http://en.wikipedia.org/wiki/List_of_programmers

Anonymous Sun Feb 20 03:16:57 2011 No.2570604

The game with Watson highlighted why Jeopardy is a terrible game for people who are too slow on the buzzer:
Even if you know the correct answer, if you do not respond FIRST you will lose the game.

It would have been a much more serious game if either human player responded to the question anyways before Alex says who's right or not, and see who gets what correct.

My bet is, both humans would show they know and score more then what they did in the OP. Probably not beat Watson but still, much closer then what it ended up being.

>>	Anonymous Sun Feb 20 03:17:34 2011 No.2570605 >>2570588 dumb as fuck

>>	Anonymous Sun Feb 20 03:18:57 2011 No.2570610 >>2570605 dumber than fuck

>>	Anonymous Sun Feb 20 03:19:34 2011 No.2570616 >>2570588 if it was fed a scanned image right? that is still cheating ass-hat it needs to READ the question the same way the humans do, i.e. processing optical (or audio) input

Anonymous Sun Feb 20 03:20:57 2011 No.2570619

>>2570604
That's not how you play Jeopardy! Ken won 70-howevermany games not because he knew more, but because he was superfast at guessing whether or not he would be able to think of the answer, and because he sat in his hotel room practicing buzzing in every night.

>>	Anonymous Sun Feb 20 03:21:34 2011 No.2570623 feeding files to a machine is not playing JEOPARDY! Watson was fed files. When it can READ the answer or LISTEN to the answer, then it will be playing JEOPARDY! Watson was a MARKETING HOAX

>>	Anonymous Sun Feb 20 03:22:34 2011 No.2570629 >>2570623 what watson dis was the equivalent of saying a machine can play baseball, and instead of actually hitting a pitch, it just is fed a ball, which it shoots at the defending team DERP robots kill us at baseball, fear dur singularitard

>>	Anonymous Sun Feb 20 03:29:34 2011 No.2570654 IF YOU STILL THINK WATSON WAS IMPRESSIVE READ THESE: >>2570623 >>2570629

>>	Scientist Sun Feb 20 03:31:57 2011 No.2570664 >>2570629 Not really. Parsing spoken human language into text is the easy part. We already have programs which do that with pretty good degrees of success. The ridiculously hard part is getting all of the subtle nuance, which Watson does quite well.

Anonymous Sun Feb 20 03:32:57 2011 No.2570667

>>2566040
The people who are "actually developing AI" are a bunch of nerds mentally masturbating over mathematics. This system can emulate any guy with access to google answering Jeopardy questions. It's Jeopardy-intelligent.

That's true of any human. Humans are usually bad at most things, except one or two.

Anonymous Sun Feb 20 03:35:57 2011 No.2570687

Of course it's an amazing computer, It's a fucking internet google with a buzzer attached, of course it fucking won. Fair would be pitting it against another computer, with the same buzzer mechanism, as in it's down to the computer system itself.

Why the fuck would anyone watch the show and expect anything different to happen, heard that on the radio and figured anyone with the littlest computer literacy couldn't possibly have any interest in watching.

Anonymous Sun Feb 20 03:35:57 2011 No.2570688
File: 5 KB, 253x300, ]ass.jpg [View same] [iqdb] [saucenao] [google]

>>2570623
your conspiracy theory needs at least some twisted fact to use as proof, don't you tinfoil hats know anything?

>>2570629
holy shit you're a retard, no really, you've got brain damage.

>>	Anonymous Sun Feb 20 03:36:34 2011 No.2570692 >>2570664 wrong. the hard part is getting the answer before the trebek shuts his mouth so you know to buzz or not Watson had a HUGE advantage by getting fed a text file

>>	Anonymous Sun Feb 20 03:37:34 2011 No.2570703 IF YOU STILL THINK THE WATSON HOAX WAS IMPRESSIVE READ THESE: >>2570623 >>2570629

>>	Anonymous Sun Feb 20 03:38:57 2011 No.2570712 >>2570692 No, everyone gets this, and everyone who knows the answer has buzzed in long before trebek closes his mouth.

>>	Scientist Sun Feb 20 03:40:57 2011 No.2570721 >>2570692 You are aware the text file was given to him only after the whole spoken question was done? You are aware that allowing Watson to receive the question at the same rate as the contestants could only increase his competitive advantage?

>>	Anonymous Sun Feb 20 03:41:57 2011 No.2570726 >>2570687 Well, then you're retarded. Just a year ago, the system was getting creamed by human opponents. Even in this match it was a tie after the jeopardy round. If he was a little less conservative in his final jeopardy wager, he would have lost.

>>	Anonymous Sun Feb 20 03:44:34 2011 No.2570749 IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE READ THESE: >>2570623 >>2570629

>>	Anonymous Sun Feb 20 03:45:34 2011 No.2570762 >>2570721 yea thats a lie, Watson got the text file as soon as the answer was displayed on the screen

Anonymous Sun Feb 20 03:46:57 2011 No.2570767

>>2565741
>>the humans repeatedly press the button
>Watson still beats them because he can just spam the electrical impulse for buzzing in
>people are amazed it won

KEN JENNINGS on The Buzzer:
"As silly as it is that Watson could out-buzz two very good buzzers nearly at will, it would be MUCH MORE SILLY to build in some kind of handicap to slow it down to human reflex speeds. Does that really seem like a "fairer" fight to anyone? Forcibly prevent one contestant, Harrison Bergeron-style, from using its greatest strength? Maybe Deep Blue should have been programmed with the most sophisticated chess strategy possible, but only been able to look ahead 3 to 5 moves, since a human player can't consistently look ahead any more than that. Screw that--it's a digital player, it should be able to do what digital machines are good at. Like buzzer-timing."

tl;dr
Fuck off OP.

>>	Anonymous Sun Feb 20 03:46:57 2011 No.2570771 >>2570749 LOL, fuck of retard.

>>	Scientist Sun Feb 20 03:47:57 2011 No.2570779 >>2570762 Orly? Nevermind then.

>>	Anonymous Sun Feb 20 03:48:57 2011 No.2570781 File: 31 KB, 349x642, retard.jpg [View same] [iqdb] [saucenao] [google] >>2570749

>>	Anonymous Sun Feb 20 03:49:57 2011 No.2570787 >>2570762 See; >>2570767 Even Ken Jennings thinks you're a fucking retard. Why hinder the computer from doing... Exactly what it's good at, and is suppose to be doing...

>>	Anonymous Sun Feb 20 03:49:57 2011 No.2570788 File: 28 KB, 500x500, suck-it-trebek.jpg [View same] [iqdb] [saucenao] [google] I've just completed Watson 2.0. Pic related.

>>	Anonymous Sun Feb 20 03:50:34 2011 No.2570797 IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE READ THESE: >>2570623 >>2570629 Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX.

>>	Anonymous Sun Feb 20 03:50:57 2011 No.2570798 proof there was a text file even used put up or shut up, dipshits

>>	Anonymous Sun Feb 20 03:52:34 2011 No.2570812 >>2570798 PROOF of any rules? IBM HIDES THE RULES. This is proof of MARKETING HOAX

>>	Anonymous Sun Feb 20 03:53:57 2011 No.2570817 >>2570797 Wait, what exactly are you trying to say? Watson is a search engine and not really AI? That's exactly what it's suppose to be you moron.

Anonymous Sun Feb 20 03:55:57 2011 No.2570835

>>2570812
Which part is the hoax, the fact that he "won"?

It doesn't change the fact that IBM is still incredibly competent when it comes to technological developments. Why in the world it matters whether or not they paid a couple of Jeopardy participants to supposedly throw an obviously pre-meditated game, I'll never understand.

>>	Anonymous Sun Feb 20 03:56:57 2011 No.2570838 File: 50 KB, 250x207, 1243281776748.jpg [View same] [iqdb] [saucenao] [google] >>2570812 so your proof is that you cant find any evidence? you have fun with that one.

>>	Anonymous Sun Feb 20 03:57:34 2011 No.2570845 >>2570798 http://www.youtube.com/watch?v=ZLdkJpAtt1I 2:20 seconds fed the answer electronically

Anonymous Sun Feb 20 03:58:34 2011 No.2570853

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or some kind of "achievment"

READ THESE:
>>2570623
>>2570629

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX.

>>	Anonymous Sun Feb 20 04:00:57 2011 No.2570862 >>2570853 IF YOU'RE STILL POSTING THIS SHIT, BECAUSE YOU'RE AN ILLITERATE JACKASS READ THESE: >>2570817 >>2570767 Watson did exactly what it was suppose to do.

>>	Anonymous Sun Feb 20 04:05:57 2011 No.2570886 >>2570853 The fact that people like you exist makes me wish I could leave this planet.

Anonymous Sun Feb 20 04:06:34 2011 No.2570892

>>2570862

Watson had an unfair advatage, not in using the buzzer, but in being able to start processing the search information instantly, instead of having to interpret audio/visual information LIKE THE HUMANS.

THE HUMANS HAD MORE PROCESSING TO DO.

IT WAS A MARKETING HOAX.

WATSON DID NOT PLAY JEOPARDY!

Anonymous Sun Feb 20 04:07:34 2011 No.2570897

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or some kind of "achievment"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

Anonymous Sun Feb 20 04:07:57 2011 No.2570898

>>2570892
>Watson had an unfair advatage, not in using the buzzer, but in being able to start processing the search information instantly, instead of having to interpret audio/visual information LIKE THE HUMANS.

So, what you're trying to say is:
The machine didn't have an unfair advantage because it was able to buzz like a machine, but because it was able to process like a machine.

You're a retard, fuck off.

>>	Anonymous Sun Feb 20 04:08:57 2011 No.2570905 >>2570892 Except that humans have an evolutionary advantage of a couple billion years. The very least you would expect is they can interpret human-made signals with ease.

>>	Anonymous Sun Feb 20 04:08:57 2011 No.2570907 >>2570897 IF YOU'RE STILL POSTING THIS SHIT, BECAUSE YOU'RE AN ILLITERATE JACKASS READ THESE: >>2570817 >>2570767 AND NOW APPARENTLY THIS: >>2570787 Watson did exactly what it was suppose to do.

>>	Anonymous Sun Feb 20 04:09:34 2011 No.2570913 >>2570898 bullshit, machines can listen to audio, machines can read text with scanners. WATSON DID NOT LISTEN TO AUDIO. WATSON DID NOT READ THE VISUAL. HUMANS HAD FAR MORE PROCESSING. WATSON DID NOT PLAY JEOPARDY!

Anonymous Sun Feb 20 04:10:34 2011 No.2570918

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or some kind of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

>>	Anonymous Sun Feb 20 04:12:57 2011 No.2570926 >>2570913 Holy shit, you can not get more retarded. See; >>2570898 And >>2570907

>>	Anonymous Sun Feb 20 04:12:57 2011 No.2570927 >>2570918 IF YOU'RE STILL POSTING THIS SHIT, BECAUSE YOU'RE AN ILLITERATE JACKASS READ THESE: >>2570817 >>2570767 AND NOW APPARENTLY THIS: >>2570787 Watson did exactly what it was suppose to do.

>>	Anonymous Sun Feb 20 04:13:57 2011 No.2570929 >>2565955 If an answer is transmitted to Watson it is like a human being given a piece of paper. You still have to read it and understand it, which takes time.

>>	Anonymous Sun Feb 20 04:14:34 2011 No.2570934 >>2570926 answer this ass-hat>>2570913 or you're IBM VIRAL MARKETER

>>	Anonymous Sun Feb 20 04:15:34 2011 No.2570944 >>2570929 bullshit, its like a human having the answer wired directly into its cerebral cortex. the machine could have read, it could have listened, but IT WOULD HAVE LOST HORRIBLY. Watson was an ADVERTISING HOAX.

>>	Anonymous Sun Feb 20 04:15:57 2011 No.2570941 File: 15 KB, 300x300, 010109010206010312200709304a0b4bf3906b76df4d00285c.jpg [View same] [iqdb] [saucenao] [google] >>2570913 >>2570897

>>	Anonymous Sun Feb 20 04:16:34 2011 No.2570951 >>2570941 IBM VIRAL MARKETING

Anonymous Sun Feb 20 04:16:57 2011 No.2570948

>>2570934
Did you just completely ignore my post?

What exactly are you having trouble understanding? Ask me what exactly you want to know, and I'll answer your question. I'm eager to, so please, just ask so I can clear up your obvious misunderstandings...

>>	Anonymous Sun Feb 20 04:17:34 2011 No.2570959 >>2570948 you're full of shit answer this ass-hat>>2570913 or you're IBM VIRAL MARKETER

>>	Anonymous Sun Feb 20 04:17:57 2011 No.2570952 >>2570944 So you're problem is with the machine not buzzing like a machine, but rather, processing information like a machine? I don't understand...

Anonymous Sun Feb 20 04:17:57 2011 No.2570953

>>2570623
You are confusing being handed a piece of paper with words on it with reading and understanding those words.

You may instantly receive a message by watching it on a video screen or being handed a piece of paper. But you still have to read it and understand it.

Everybody received the clue at the same time. For the humans it appeared on their retina almost instantly when the clue was revealed.

Just because Watson got an electronic message containing the clue is no advantage. He still has to read it. Just like you have to read this post.

>>	Scientist Sun Feb 20 04:17:57 2011 No.2570956 >>2570944 I think you grossly overestimate the processing time required for a computer to change spoken word into text. It's unmeasurable for the purposes of Jeopardy vs humans.

Anonymous Sun Feb 20 04:19:34 2011 No.2570971

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or was some kind of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

>>	Anonymous Sun Feb 20 04:19:34 2011 No.2570973 File: 66 KB, 423x444, 1281254936345.jpg [View same] [iqdb] [saucenao] [google] >>2570951

Anonymous Sun Feb 20 04:19:57 2011 No.2570972

>>2570944
There is a huge difference between the representation of a question and how that question is when it is inside your head.

If I transmitted a question into your brain you would still have to figure out what the words meant! You'd have to read them out of your brain.

Your understanding of words takes time, so if I just popped some words into your mind you are no better of than if I flashed them on a computer video screen!

In other words, I'm transmitting text directly into your brain right now!

Anonymous Sun Feb 20 04:20:57 2011 No.2570976

>>2570959
Your ignorance is showing, so here, I'll ask the question for you.

>"Did Watson have an unfair advantage"
In what sense? Being a machine?

>"No, because he didn't have to listen or read"
He also didn't have to physically press a buzzer, but you have no problem with this... Alas;
He 'read'. Being given data doesn't give you an instantanous answer, he still has to process that data, in this case the question, in much the same way a Human would read it, or see it.
Does a Human who can read quicker have an unfair advantage over one who reads slower? Would the competition be brought into question if such a scenario arose?

Anything else?

>>	Anonymous Sun Feb 20 04:21:57 2011 No.2570977 What is wrong with IBM advertising?

>>	Anonymous Sun Feb 20 04:21:57 2011 No.2570978 File: 55 KB, 450x326, 1284042949391.jpg [View same] [iqdb] [saucenao] [google] >>2570959

>>	Anonymous Sun Feb 20 04:21:57 2011 No.2570980 >>2570976 >instantaneous Apologies, I wrote that in a fit of "Holy shit, this guy is a retard".

>>	Anonymous Sun Feb 20 04:22:34 2011 No.2570983 >>2570972 you're pretending it take ZERO seconds to read a question. It takes time AND IT TAKES PROCESSING POWER. HUMANS DID MORE PROCESSING, which Watson could have done, but was exempted from. Watson was an ADVERTISING HOAX.

>>	Anonymous Sun Feb 20 04:22:57 2011 No.2570984 >>2570976 He did physically press a buzzer.

>>	Anonymous Sun Feb 20 04:23:57 2011 No.2570988 >>2570984 He spams the button electronically, he doesn't have a robot arm with various muscles and such - or am I mistaken?

>>	Anonymous Sun Feb 20 04:24:57 2011 No.2570993 >>2570983 THE OPPONENT READ FASTER THAN I DID, THAT MEANS THIS COMPETITION IS A FRAUD!!!!!! Honestly, pick yourself up, get the fuck out of the room. Turn the computer off before you leave.

Scientist Sun Feb 20 04:24:57 2011 No.2570996

>>2570983
CS fag here. Again, not enough time to matter. The time taken to understand the question in text form, and search for an answer, easily dwarfs the time it would take to take the spoken question and convert it to text. The difference is so huge this is a pedantic and rather useless point to argue.

Anonymous Sun Feb 20 04:25:57 2011 No.2570997

>>2570983
No, you are the one who believes it takes zero seconds for Watson to read a question because it was transmitted in electronic form.

Watson still has to read it. It is sent to him raw and unprocessed. This is exactly analogous to being presented with words on a screen. Before you read them they are just a pattern on your retina.

Before watson starts to parse the question, they are just a meaningless pattern of bits.

Also, for humans, reading and understanding go together. You do not read at one point and then start understanding at another. For humans, reading and answering a question happen in parallel.

Same for Watson.

You are an idiot. If you don't get this I'm not trying to help you understand it anymore.

>>	Anonymous Sun Feb 20 04:25:57 2011 No.2570998 >>2570983 It takes processing power to a simpleton like you. Most humans do it without the need to think. Otherwise it would imply that unicellular organisms equipped with light-sensitive patches are intelligent.

Anonymous Sun Feb 20 04:26:34 2011 No.2571001

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or was some kind of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could have done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

>>	Anonymous Sun Feb 20 04:26:57 2011 No.2571005 File: 119 KB, 319x480, 1279419164322.jpg [View same] [iqdb] [saucenao] [google] >>2570983

>>	Anonymous Sun Feb 20 04:27:57 2011 No.2571011 File: 94 KB, 510x680, 1243856547150.jpg [View same] [iqdb] [saucenao] [google] >>2571005

Anonymous Sun Feb 20 04:28:57 2011 No.2571014

>>2570996
Yes, I agree, if you were to throw as much processing power as Watson has at speech recognition then they could probably add a couple of milliseconds to the total time it takes watson to answer.

Next year, due to Moorse law, the time it takes him to answer will be about HALF. So if he doesn't win this year he will next.

After 15 years do you really think a human could win at chess? In 15 years Watson will fit in a laptop and be integrated into all the search engines. In 30 it will be in your cell phone.

At that time do you really think there will be room to argue if Watson is better at Jeopardy than humans? It is stupid and pointless to argue since if it is close now, it won't even be close at all very soon.

>>	Anonymous Sun Feb 20 04:29:57 2011 No.2571016 File: 23 KB, 239x244, 1272832005331.jpg [View same] [iqdb] [saucenao] [google] >>2571011

>>	Anonymous Sun Feb 20 04:30:57 2011 No.2571022 >>2571001 Could you post this a few thousand more times? I don't think you made your point.

>>	Anonymous Sun Feb 20 04:30:57 2011 No.2571025 File: 17 KB, 283x420, 1224318717474.jpg [View same] [iqdb] [saucenao] [google] >>2571016

>>	Scientist Sun Feb 20 04:30:57 2011 No.2571026 >>2571014 >milliseconds That's being very very generous.

>>	Anonymous Sun Feb 20 04:31:57 2011 No.2571027 >>2571001 IF YOU'RE STILL POSTING THIS SHIT, BECAUSE YOU'RE AN ILLITERATE JACKASS READ THESE: >>2570817 >>2570767 AND NOW APPARENTLY THESE: >>2570787 >>2570976 Watson did exactly what it was suppose to do. It acted like a machine.

Anonymous Sun Feb 20 04:32:34 2011 No.2571033

>>2571014

sppech recognition is bullshit, because then Watson would be WAY behind in terms of processing time before the buzzers open.

if it wants to compete on equal footing, it needs a camera pointed at the screen and reading the question, just like the humans do.

and if it's sooo close like you singulari-tards believe, then why couldn't IBM just wait a couple of months and have an honest competition?

Watson was an ADVERTISING HOAX.

Anonymous Sun Feb 20 04:33:34 2011 No.2571038

IF YOU STILL THINK THE WATSON-JEOPARDY! HOAX WAS IMPRESSIVE or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could have done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

>>	Anonymous Sun Feb 20 04:33:57 2011 No.2571036 File: 513 KB, 1196x1080, 1273945789859.jpg [View same] [iqdb] [saucenao] [google] >>2571001

>>	Anonymous Sun Feb 20 04:35:34 2011 No.2571050 Watson was a HOAX! case in point:>>2570983 Watson was PAID ADVERTISING! NOT AN HONEST GAME OF JEOPARDY!

Scientist Sun Feb 20 04:35:57 2011 No.2571051

>>2571033
I think the singularly talk is nonsense. I also think that speech and image recognization is trivial in processing power and time compared to understanding the question and finding the answer. If you required Watson to have a camera to parse the question, or listen to the question to parse it, that would add negligible time to his response time.

>>	Anonymous Sun Feb 20 04:36:57 2011 No.2571052 File: 363 KB, 853x1023, 1281253556377.jpg [View same] [iqdb] [saucenao] [google] >>2571033

Anonymous Sun Feb 20 04:37:57 2011 No.2571056

>>2571026
Yeah, everything in my post was very conservative. Even being conservative it just seems silly to argue against IBM's achievement.

Not only that, it completely ignores that Watson is impressive without being put into a competitive situation!

Category: Presidential Rhymes
Clue: Barracks Pack Animals
Question: What are Obama's Llamas

How the fuck??

Category: Northern Most Capitols
Clue: Paris, Dublin, Stockholm
Question: What is Stockhom

Double WTF?? I mean, did they specifically add a heuristic for geographical location of cities? Because you are never going to find the phrase "Of the cities Stockholm, Dublin, and Paris, Stockholm is the most north" in ANY database.

It requires putting together a great many different things to arrive at that answer.

I'm just blown away.

Does it even matter if it was competing with anybody?

Anonymous Sun Feb 20 04:39:34 2011 No.2571063

If you still believe the Watson-JEOPARDY! HOAX WAS IMPRESSIVE or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could have done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

>>	Anonymous Sun Feb 20 04:39:57 2011 No.2571065 File: 114 KB, 416x428, UMAD.png [View same] [iqdb] [saucenao] [google] >>2571050

Anonymous Sun Feb 20 04:40:57 2011 No.2571069

>>2571050
Nobody is denying that Watson is done as publicity. That doesn't change anything about what Watson is or what it can do.

If all you are arguing is that Watson can buzz fast and doesn't have to read/listen then so what?

The important thing is that it can answer questions.

Unless you have something that shows Watson did not answer those questions your arguments are pointless and silly.

>>	Anonymous Sun Feb 20 04:41:34 2011 No.2571070 >>2571056 has anybody put those answers into google? show me the first google answer to those is fucking irrelevant, then i'll be impressed. because watson vs. google would have been an HONEST COMPETITION. watson-JEOPARDY! was a HOAX.

>>	Anonymous Sun Feb 20 04:42:34 2011 No.2571072 >>2571069 bullshit, it was billed as a fair competition to the american public. IT WAS A HOAX.

>>	Anonymous Sun Feb 20 04:43:57 2011 No.2571073 File: 14 KB, 300x202, sears.jpg [View same] [iqdb] [saucenao] [google] >>2571063

>>	Anonymous Sun Feb 20 04:44:57 2011 No.2571075 >>2571072 It was a public display of abilities. Nobody ever said that he was going to play identically to a human.

>>	Anonymous Sun Feb 20 04:45:34 2011 No.2571080 Watson was an ADVERTISING HOAX! case in point:>>2570983 Watson was PAID ADVERTISING! NOT AN HONEST GAME OF JEOPARDY!

>>	Anonymous Sun Feb 20 04:45:57 2011 No.2571081 File: 9 KB, 288x371, 1272598555608.jpg [View same] [iqdb] [saucenao] [google] >>2571072

Anonymous Sun Feb 20 04:46:57 2011 No.2571084

>>2571056
Yes, they have geospatial reasoning built in. They have to be able to tell, in an example they gave if the question is implying something about landing in india, and a text has something about arriving on a particular beach, and that beach happens to be in india, the logic has to draw those things together. Similarly to how it does the same thing with words.

This troll
>>2571070
>>2571063
should be ignored. He's a fucking idiot.

>>	Anonymous Sun Feb 20 04:46:57 2011 No.2571085 File: 17 KB, 460x288, 1274485530471.jpg [View same] [iqdb] [saucenao] [google] >>2571080

>>	Anonymous Sun Feb 20 04:47:34 2011 No.2571087 >>2571075 bullshit, it was billed as a fair competition to the american public. IT WAS A HOAX. HUMANS DID MORE PROCESSING see>>2570983

Anonymous Sun Feb 20 04:48:34 2011 No.2571093

If you still believe the Watson JEOPARDY! HOAX WAS IMPRESSIVE or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could have done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

Anonymous Sun Feb 20 04:48:57 2011 No.2571094

>>2571072
I think you must be mad about something.

It was billed for what it was, two humans vs. a machine.

If the machine could not answer the questions correctly everything you are arguing against would be pointless because nothing would keep it from buzzing in and scoring -100k

>>	Anonymous Sun Feb 20 04:50:57 2011 No.2571100 >>2571087 Humans did more processing? Was this a contest of how much processing was happening? No, Jeopardy is not a processing contest. It is a trivia game show. Are you retarded? I guess cars are slower than humans because they do less running.

>>	Anonymous Sun Feb 20 04:53:34 2011 No.2571109 >>2571100 if a car and a human raced, they would both run the same distance, Watson ran a shorter distance, because it didn't have as much work to do because it was EXEMPTED FROM READING THE QUESTION.

Anonymous Sun Feb 20 04:54:57 2011 No.2571115

The game was honest. We all know how the game was played and can make up our own minds about it. We do not need to be spammed HOAX with little to no evidence that it was anything but what it was claimed.

We do not disagree about anything here actually. We both know the rules. We both know how Watson got its information. We know how it buzzed in. We agree it answered the questions without assistance from humans.

So, what you are saying is like YOUR FUCKING OPINION.

It isn't some conspiracy to sell IBM stock. We even agree it was an ad for IBM. Like that makes a difference.

So... What is your problem?

You want us to stop liking stuff you don't like?

>>	Anonymous Sun Feb 20 04:55:57 2011 No.2571119 >>2571109 Watson had to read a question. Unless you think receiving a message is the same as reading a message. If Watson didn't read the message then how did he answer the question? I'd like to see you answer a question without reading it.

Anonymous Sun Feb 20 04:56:34 2011 No.2571124

>>2571094

bullshit, it was billed as a fair competition on a well known game show. it was, in fact, nothing more than PAID ADVERTISING.

Watson did not play JEOPARDY!

see>>2570983

Watson was given a large speed advantage proportional to the short time before buzzing in

MR. RAGE !D9l9S8Lio6 Sun Feb 20 04:56:57 2011 No.2571125

>>2571109

TO CONTINUE THE METAPHOR, WATSON RAN A DISTANCE ~2" SHORTER OVER A MILE-LONG RACE.

HEAT-WARPING OF THE TRACK WOULD MAKE A BIGGER DIFFERENCE THAN 2".

THE REASON THOSE 2" WEREN'T ADDED IS BECAUSE THEY WERE IRRELEVANT TO THE DISPLAY. IT WOULD'VE BEEN A POINTLESS ADDITIONAL GIMMICK.

THE POINT, FOR THE RECORD, WAS THAT WATSON COULD ANSWER QUESTIONS CORRECTLY *AT ALL*. COULD THEY HAVE DONE THE SAME TEST BY SIMPLY HAVING A SCIENTIST READ QUESTIONS TO WATSON IN A LABORATORY? SURE. BUT THAT WOULDN'T HAVE GOTTEN AS MUCH PUBLICITY OR MONEY.

THE "INTEGRITY" OF THE TEST IS THE SAME IN BOTH CASES. YOU'RE BITCHING THAT SOMETHING THAT WASN'T BEING TESTED WASN'T TESTED, ESSENTIALLY. IN WHICH CASE "FUCKING DUH."

>>	Anonymous Sun Feb 20 04:56:57 2011 No.2571127 >>2571109 oh youre stupid watson didnt run hes a computer

>>	Scientist Sun Feb 20 04:57:57 2011 No.2571135 >>2571109 Which is like 1 mm on a 10 mile race. Congrats.

Anonymous Sun Feb 20 04:59:34 2011 No.2571142

If you still believe the Watson-JEOPARDY! HOAX was "impressive" or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could have done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

Anonymous Sun Feb 20 04:59:57 2011 No.2571141

>>2571124
How did Watson not play Jeopardy?

If you are arguing that the rules were not followed exactly. Everybody agreed to the new rules and everybody knows the new rules.

Also, the rules of Jeopardy have changed slightly since the 80s. I guess nobody nowadays is really playing it.

>>	Anonymous Sun Feb 20 04:59:57 2011 No.2571143 File: 28 KB, 200x279, AmazingSpiderman215-03.jpg [View same] [iqdb] [saucenao] [google] Shut up you idiot troll. You're not convincing anybody of your position, and we all know you're just trying to bait responses. Everybody stop replying him.

>>	Anonymous Sun Feb 20 05:00:34 2011 No.2571155 >>2571135 bullshit, the buzzing in comes down to milliseconds which is very near the time advantage Watson was given

Anonymous Sun Feb 20 05:00:57 2011 No.2571153

So what is it all good for?

Give a computer a huge assortment of texts in a certain field, say physics.
Ask it a question.
Get the answer.

2nd step: Let it ask itself a question.
Generate new text.
Ask new question.
Rinse and repeat.

The singularity. It is almost here.

>>	Anonymous Sun Feb 20 05:01:34 2011 No.2571161 >>2571141 the rules are the same for all contestants in any given match. EXCEPT for Watson, who was given a large advantage by being EXEMPTED FROM READING THE QUESTION.

>>	Anonymous Sun Feb 20 05:01:57 2011 No.2571159 >>2571153 But I do not think Watson knows what questions to ask. So still a ways to go

>>	Anonymous Sun Feb 20 05:02:57 2011 No.2571162 > people think reading and pressing buttons is the hardest part of Jeopardy You should donate your higher brain functions to Watson, he'll make much better use of it.

>>	Scientist Sun Feb 20 05:02:57 2011 No.2571167 >>2571153 >2nd step: Let it ask itself a question. That's the hard part. Singularity is still far off.

>>	Scientist Sun Feb 20 05:03:57 2011 No.2571172 >>2571155 And we're talking microseconds (or better) for it to convert speech to text.

>>	Anonymous Sun Feb 20 05:03:57 2011 No.2571176 >>2571155 a millimeter compared to a mile is much smaller than a millisecond compared to 5 seconds Your argument boils down to this: Watson is not a human being so he cannot play Jeopardy.

>>	Anonymous Sun Feb 20 05:04:57 2011 No.2571181 >>2571161 How can you answer a question without reading it?

Anonymous Sun Feb 20 05:05:34 2011 No.2571188

>>2571176

wrong

Watson could have taken a picture of the screen, scanned the picture to read the text, then input the text. Just like a human.

Instead, it skipped all those steps and went straight to input the text.

Watson was given a huge advantage.

stop PRETENDING it was a FAIR competition.

>>	Anonymous Sun Feb 20 05:05:57 2011 No.2571190 >>2571162 There really is absolutely nothing impressive about Watson. Had it been the same thing with speech recognition it would've been an amazing step forward technologically speaking. It has nothing to do with Jeopardy.

>>	Anonymous Sun Feb 20 05:06:57 2011 No.2571192 >>2571161 You can have a game where each player has different rules. Everybody agreed before hand. So what was unfair?

>>	Scientist Sun Feb 20 05:06:57 2011 No.2571193 >>2571190 You are a retard.

>>	Anonymous Sun Feb 20 05:07:34 2011 No.2571204 >>2571192 It was billed to the american public as a FAIR COMPETITION. Watson was given a HUGE ADVANTAGE. WATSON_JEOPARDY! WAS A HOAX.

>>	Anonymous Sun Feb 20 05:07:57 2011 No.2571199 >>2571190 Speech to Text is easy. Your cell-phone can do it. How is that impressive?

Anonymous Sun Feb 20 05:09:34 2011 No.2571211

>>2571176

WRONG

Watson could've taken a picture of the screen, scanned the picture to read the text, then input the text. Just like a human.

Instead, it skipped all those steps and went straight to input the text.

Watson was given a huge advantage.

stop PRETENDING it was a FAIR competition.

Anonymous Sun Feb 20 05:09:57 2011 No.2571210

>>2571188
If everybody agrees to something it is fair.

Taking a picture of the screen and recognizing the text is just as easy as receiving the clue directly.

There is an IPHONE APP that reads text like this. It is easy.

Again, how do you answer a question without reading it?

>>	Scientist Sun Feb 20 05:09:57 2011 No.2571212 >Watson was given a HUGE ADVANTAGE. He had the advantage of 1 mm on a mile race. It's negligible.

>>	MR. RAGE !D9l9S8Lio6 Sun Feb 20 05:09:57 2011 No.2571215 >>2571204 >huge >literally less than .1% of an advantage YOU ARE, AGAIN, TALKING ABOUT THINGS THAT HAVE A SMALLER IMPACT ON THE CONTEST THAN SIMPLE ENVIRONMENTAL FACTORS AND RANDOM FLUCTUATIONS.

>>	Anonymous Sun Feb 20 05:10:34 2011 No.2571218 >>2571212 bullshit see this>>2571211 that is not a millisecond of processing time

>>	Anonymous Sun Feb 20 05:10:57 2011 No.2571217 >>2571211 If everybody agrees to something it is fair. You cannot answer a question unless you read it first.

>>	Anonymous Sun Feb 20 05:11:57 2011 No.2571223 >>2571218 It would take nanoseconds to read text out of a video image. Basically the exact same time it would take to transmit it in ASCII code.

>>	Anonymous Sun Feb 20 05:11:57 2011 No.2571224 >>2571159 True. But does it have to? It could just brute-force science: Ask ANY question possible.

>>	Anonymous Sun Feb 20 05:11:57 2011 No.2571225 >>2571211 I guess the problem is that no one working on the project was a FUCKING RETARD like you, so they just did it the simpler way, with a text file instead of using COMPETELY UNNECESSARY OCR scanning.

>>	Scientist Sun Feb 20 05:11:57 2011 No.2571226 >>2571218 You're right. It's much less than a millisecond of processing time.

Anonymous Sun Feb 20 05:12:34 2011 No.2571227

>>2571215

look bro,

Watson could have taken a picture of the screen, scanned the picture to read the text, then input the text. Just like a human.

Instead, it skipped all those steps and went straight to input the text.

Watson was given a huge advantage.

stop PRETENDING it was a FAIR competition.

THIS IS FAR MORE THAN A "MILLISECOND" or ".1%" ADVANTAGE.

Anonymous Sun Feb 20 05:12:57 2011 No.2571232

Yes, it has to ask the right questions.

If all you had to do was randomly generate questions then we would have computer proofs for all mathematical theorems at our fingertips. But we don't because the universe would die before the computer finished proving the simplest things.

>>	Anonymous Sun Feb 20 05:13:57 2011 No.2571234 >>2571199 If it's so easy why didn't they do it?

>>	Scientist Sun Feb 20 05:14:57 2011 No.2571240 >>2571227 No man, on the kind of hardware they had, it's less than a millisecond. Much less.

>>	Anonymous Sun Feb 20 05:14:57 2011 No.2571242 >>2571227 OCR of video image takes nanoseconds Speech recognition takes microseconds Neither of these things requires any intelligence. Both of these things can be done by a toaster oven. How do you answer a question without reading it first?

>>	Anonymous Sun Feb 20 05:15:34 2011 No.2571245 >>2571234 obviously, this is the hoax, several of the people arguing for watson are IBM VIRAL MARKETERS, tho i'm sure more than a few comp sci and physi majors are getting the hero disillusioned too.

MR. RAGE !D9l9S8Lio6 Sun Feb 20 05:15:57 2011 No.2571243

>>2571234

BECAUSE IT WOULD'VE COST MORE AND BEEN A PAIN IN THE ASS TO SET UP, FOR AN UTTERLY TRIVIAL DETAIL IRRELEVANT TO THE ACTUAL TESTS BEING CONDUCTED?

YOU'RE ASKING SOMETHING EQUIVALENT TO "WELL IF IT'S SO EASY FOR ALL THE SCIENTISTS TO WEAR BLUE, WHY DIDN'T THEY?" AND BITCHING ABOUT THE RESULTS OF A PHYSICS EXPERIMENT BECAUSE ONE SCIENTIST WAS WEARING RED.

>>	Anonymous Sun Feb 20 05:16:34 2011 No.2571254 >>2571242 bullshit on your nanoseconds it would have to pay close attention to when exactly to snap the frame, it wouldn't get the frame instantly, thats for sure

>>	Anonymous Sun Feb 20 05:16:57 2011 No.2571253 >>2571234 It is pretty easy to do a lot of unnecessary things. Why don't you go do them?

>>	Anonymous Sun Feb 20 05:16:57 2011 No.2571255 >>2571227 No its not. For a camera to read the question and covert it to THE SAME DATA Watson was given would take LESS THAN A MILLISECOND and would have impact on the game.

>>	MR. RAGE !D9l9S8Lio6 Sun Feb 20 05:17:57 2011 No.2571262 >>2571254 >snap the frame >not, y'know, a fucking video camera like the human eye is THE FUCK IS WRONG WITH YOU?

>>	Anonymous Sun Feb 20 05:17:57 2011 No.2571263 >>2571245 Be sure to tell IBM where to send my cheque.

>>	Scientist Sun Feb 20 05:18:57 2011 No.2571266 >>2571254 >pay close attention to when to snap the frame >pay close attention (via a camera) to when to use a camera >full retard

Anonymous Sun Feb 20 05:19:34 2011 No.2571272

If you still believe the Watson-JEOPARDY! HOAX was "impressive" or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could've done it, but it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

Anonymous Sun Feb 20 05:19:57 2011 No.2571274

>>2571255
You mean no impact. Recognizing text in a video image takes practically no time.

And even then, this is confusing processing a video image with "reading".

Reading something requires processing it. It does not involve just receiving it.

If Watson didn't read he could not be able to answer the questions.

Anonymous Sun Feb 20 05:21:34 2011 No.2571280

look bro,

Watson could have taken a picture of the screen, scanned the picture to read the text, then input the text. Just like a human.

Instead, it skipped all those steps and went straight to input the text.

Watson was given a huge advantage.

stop PRETENDING it was a FAIR competition.

THIS IS FAR MORE THAN A "MILLISECOND" or ".1%" ADVANTAGE.

>>	Anonymous Sun Feb 20 05:21:57 2011 No.2571285 >>2571254 No attention required. This task requires so little processing time, just process all of the frames.

>>	Scientist Sun Feb 20 05:22:57 2011 No.2571286 >>2571280 What color is the sky in your world?

>>	Anonymous Sun Feb 20 05:23:34 2011 No.2571291 >>2571274 bullshit, by that "reading" logic google, AND EVER FUCKING SEARCH ENGINE EVER, have been "reading" their input go back to comp sci and tell those fags Watson was a HOAX

Anonymous Sun Feb 20 05:23:57 2011 No.2571295

The sad part is I don't think this guy is trolling, he may be dead serious. At any rate he is so vehemently against the idea that Watson competed with the same rules as any jeopardy player and won that reason isn't going to change his mind.

Please stop trying to explain it to him. You're only trolling yourselves.

>>	Anonymous Sun Feb 20 05:24:34 2011 No.2571301 >>2571286 if this was so easy, and every fucking iphone can do it, WHY DIDN"T THEY DO IT?

Anonymous Sun Feb 20 05:25:57 2011 No.2571304

>>2571280
I'm not your bro, pal.

Watson is not a demonstration of computer vision.

It is a demonstration of answering vague questions.

It succeeded at that.

Frankly, I don't care if Watson has to scan a video feed or monitor a microphone. I DON"T CARE.

So unless you want to say Watson is not impressive because the questions weren't really that hard then you've got nothing worth saying.

You vs. Watson. Write down answers to questions, no time limit. Who wins?

>>	Scientist Sun Feb 20 05:25:57 2011 No.2571311 >>2571301 I don't know.

>>	Anonymous Sun Feb 20 05:26:34 2011 No.2571312 >>2571295 answer this sage-fag:>>2571301

>>	Anonymous Sun Feb 20 05:26:57 2011 No.2571313 >>2571280 Yay! Autosage, finally

Anonymous Sun Feb 20 05:27:34 2011 No.2571318

If you still believe the Watson-JEOPARDY! HOAX was "impressive" or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could've done it, but instead it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

>>	Anonymous Sun Feb 20 05:27:57 2011 No.2571320 >>2571301 There are lots of things that are easy to do. Why don't you go do them? How impressed is everybody? Of course, you are implying that it isn't really that hard. Go download one of the hundreds of software packages that do these things!

>>	Anonymous Sun Feb 20 05:29:57 2011 No.2571327 I'm gonna close the tab and forget this thread every existed. I'll take comfort in knowing that Watson is smarter than this fucktard.

Anonymous Sun Feb 20 05:29:57 2011 No.2571328

lulz, I haven't followed the thread and couldn't find anything conclusive via a quick google search.
Could someone tell me if I got this right:
The whole achievement of Watson was the speech recognition, but the answers to the questions (that he had to recognize, this was his only task) were given to him in plaintext?

So basically all Watson had to do is understand the question to query his database.
I got that right?

If so: who gives a fuck.

Anonymous Sun Feb 20 05:30:34 2011 No.2571332

>>2571320

bullshit, the time it would have taken to process that would have totally defeated Watson, it would have been a laughing stock, and the ADVERTISING would have FAILED.

so instead they perpetrated this HOAX, and PRETENDED IT WAS A FAIR COMPETITION.

Scientist Sun Feb 20 05:31:57 2011 No.2571337

>>2571332
>
bullshit, the time it would have taken to process that would have totally defeated Watson, it would have been a laughing stock, and the ADVERTISING would have FAILED.
And I will repeat that that is simply untrue (only because we're in auto-sage mode).

>>	Anonymous Sun Feb 20 05:32:34 2011 No.2571342 >>2571328 all the american viewers who are now convinced the "singularity" is upon us because this machine is soooo fucking "smart"

MR. RAGE !D9l9S8Lio6 Sun Feb 20 05:32:57 2011 No.2571343

>>2571328

YOU'VE GOT IT COMPLETELY BACKWARDS. WATSON WAS FED THE *QUESTIONS* IN TEXT AND THE ACHIEVEMENT WAS THAT HE COULD PARSE OUT NATURAL, VAGUE HUMAN SPEECH TO KNOW *WHAT* TO QUERY.

FOR EXAMPLE: WATSON RECEIVED A TEXT FILE THAT READS "BARACK'S PACK ANIMALS" IN THE CATEGORY OF "PRESIDENTIAL RHYMES" AND PARSED OUT THAT THE ANSWER IS "WHAT ARE OBAMA'S LLAMAS"

THAT IS *FUCKING AMAZING*.

>>	Anonymous Sun Feb 20 05:34:34 2011 No.2571352 >>2571337 YOU LIE. If what you say was true, it would have taken them less than a day to implement, in fact any comp sci undergrad could have done it in a few hours. THEY WORKED ON THIS FOR 3 YEARS

>>	Anonymous Sun Feb 20 05:35:57 2011 No.2571357 >>2571328 >The whole achievement of Watson was the speech recognition, No, it wasn't.

>>	Anonymous Sun Feb 20 05:36:34 2011 No.2571365 >>2571343 Really? you think that was amazing? screw the category info, just google "BARACK'S PACK ANIMALS" THE ANSWER IS IN THE FIRST RESULT WATSON was a HOAX! It "achieved" NOTHING!

>>	MR. RAGE !D9l9S8Lio6 Sun Feb 20 05:37:57 2011 No.2571370 >>2571365 >Barack Obama's rat pack problem O RLY.

>>	Scientist Sun Feb 20 05:37:57 2011 No.2571371 >>2571365 >THE ANSWER IS IN THE FIRST RESULT Yes, but it didn't reply the whole first result. Instead, it narrowed it down to that singular phrase, which is amazing.

>>	Anonymous Sun Feb 20 05:39:34 2011 No.2571382 >>2571370 and i said it before watson v google would have been a fair competition. Watson on JEOPARDY! WAS A HOAX.

>>	Anonymous Sun Feb 20 05:40:57 2011 No.2571388 >>2571365 >screw the category info, just google "BARACK'S PACK ANIMALS" And how to get the exact words/phrases from the context? If you don't understand why that is amazing, then you need to learn what it actually takes to do it.

>>	Anonymous Sun Feb 20 05:41:34 2011 No.2571391 >>2571371 amazing? for three years of hundreds of people working on it? it can do what any college grad sitting in front of google can, and you call that amazing? I wish i was so easily impressed

Anonymous Sun Feb 20 05:43:34 2011 No.2571404

in all honesty, the truth of the situation is marking a milestone for computer scientists.

it was not the propaganda and HOAX-ERY that it was sold as.

my problem is with scaring the american people into believing in "singularity" using an unfair, rigged competition

Anonymous Sun Feb 20 05:44:57 2011 No.2571410

>>2571343
ok, so he had the questions and probably some "background knowledge" with associations, like if he hears someone talk about "animals" and "president" and he only has one question containing a noun that is an animal and a noun that is a president, he would've taken that question?

Something like this?

Why is this thread so fail? Why aren't you talking about how the AI works?
There's an AI thread on sci fucking EVERY DAY. Now there's finally some interesting input and you idiots argue with this deranged fucktard..

>>	Scientist Sun Feb 20 05:44:57 2011 No.2571405 >>2571391 Google cannot provide the answer in a single phrase. Providing a page which contains the answer, and the answer, is a fundamentally different proposition.

Anonymous Sun Feb 20 05:45:57 2011 No.2571414

>>2571391
>amazing? for three years of hundreds of people working on it? it can do what any college grad sitting in front of google can, and you call that amazing?
The fact that it was all machine that did it, yes, that is amazing.
Eventually this will come in a portable format, a little box you can ask anything, and get the correct answer, not just a whole book or page of information like google, but actual correct context, it will "know" WHAT you asked for.

>>	Anonymous Sun Feb 20 05:46:34 2011 No.2571423 >>2571410 because it wasn't actually AI to the other guys, you see what the fuck i'm talking about? right here. this person. the american people were sold a pack of lies

>>	Scientist Sun Feb 20 05:48:57 2011 No.2571432 >>2571423 AI doesn't mean what you think it means among academic circles. It's a much weaker thing. What you're talking about is strong AI.

>>	Anonymous Sun Feb 20 05:49:34 2011 No.2571438 >>2571414 do you actually believe that is better than google? google is available today on any decent cell phone...not 15 or 30 years from now please explain how this makes even the slightest difference in a person searching for information?

>>	Scientist Sun Feb 20 05:50:57 2011 No.2571441 >>2571438 Yes. Providing a succinct single phrase answer is immeasurably better than google's "return a whole bunch of possibly unrelated pages" answer. Or at least it's a shitton harder.

>>	Anonymous Sun Feb 20 05:51:57 2011 No.2571447 >>2571438 >do you actually believe that is better than google? yes, it is orders of magnitude better than google.

>>	Anonymous Sun Feb 20 05:56:34 2011 No.2571466 >>2571441 "immeasurably" better? it will still fail the hardest searches and it will be saving a few seconds on easy shit great, i'm 15 years away from saving ten seconds per google search OH NOES THE "SINGULARI-DERP"!

>>	Scientist Sun Feb 20 05:56:57 2011 No.2571469 >>2571466 I'm not arguing for a singularity you dipshit.

Anonymous Sun Feb 20 05:59:34 2011 No.2571477

>>2571469

so be honest, was their a single question that Watson got, that you couldn't get in 15 seconds on google, RIGHT NOW, TODAY?

personally the answer is no, but i am fairly skilled on google. still this is near zero actual achievement, it is an arcane hallmark for programmers

>>	Scientist Sun Feb 20 06:00:57 2011 No.2571479 >>2571477 Dunno. The key difference is that a human would be required to find it. Watson did it without human interaction, which is what makes it awesome.

>>	Anonymous Sun Feb 20 06:00:57 2011 No.2571480 >>2571272 Some of those were good points, but as already stated, this was a game show, which, I hope you know, is not held up to scientific rigor. I didn't really feel it was a huge achievement to have Watson do this either.

Anonymous Sun Feb 20 06:06:34 2011 No.2571499

>>2571480

Of course, but if you were around for the full thread, you would've seen plenty of /sci/entists that were confused by the rules and had vastly inflated notions of what was actually accomplished.

If that was true for these somewhat educated people, imagine how fucking brainwashed the average person at home was.

/sci/entists need to explain this shit in context and not just repeat(and thereby enhance) the propaganda.

>>	Anonymous Sun Feb 20 06:08:57 2011 No.2571504 >>2571499 Why would this time be different from any other moment in OMG SCIENCE media?

>>	Anonymous Sun Feb 20 06:11:34 2011 No.2571517 >>2571504 JEOPARDY! is a competition. the only thing even close to comparable was deep blue v. Kasparov. But that was a real and fair competition, not a paid advertisement.

Anonymous Sun Feb 20 06:12:34 2011 No.2571522

If you still believe Watson-JEOPARDY! HOAX was "impressive" or was some sort of "achievement"

READ THESE:
>>2570623
>>2570629
>>2570892

Jennings and the other fag were PAID to participate, unlike KASPAROV (v. deep blue). They were PAID for their role in the ADVERTISING HOAX

The "buzzer" issue is irrelevant, IBM MARKETING RED HERRING. Watson DID LESS PROCESSING than the humans. It could've done it, but instead it was exempted from it.

Watson DID NOT PLAY JEOPARDY! Watson was a MARKETING HOAX.

Anonymous Sun Feb 20 06:25:34 2011 No.2571559

Watson could've taken a picture of the screen, scanned the picture to read the text, then input the text. Just like a human.

Instead, it skipped all those steps and went straight to input the text.

Watson was given a huge advantage.

stop PRETENDING it was a FAIR competition.

>>	Anonymous Sun Feb 20 15:18:57 2011 No.2573567 >>2571559 Watson could have had a physical arm with joints, and a hand that pressed a buzzer. Stop trying to say it was fair competition, Watson had an unfair advantage of doing exactly what a machine should do.

Advanced search
Text to find
Subject [?]Search by post subject. Leave empty for any.
Username [?]Search for user name. Leave empty for any user name.
Tripcode [?]Search for tripcode. Leave empty for any.
Email [?]Search by email. Leave empty for any.
Filename [?]Search by image filename. Leave empty for any.
From Date [?]Enter what date to start searching from. Format is YYYY-MM-DD
To Date [?]Enter what date to start searching until. Format is YYYY-MM-DD
Image hash
Search in	All Posts OPs Only
Deleted posts	Show all posts Show only deleted posts Only show non-deleted posts
Internal posts	Show all posts Show only internal posts Show only archived posts
Order	New posts first Old posts first
Capcode	All Posts Only by Users Only by Mods Only by Admins Only by Developers
Results	Posts Threads
Action	[ Simple ]

/sci/ - Science & Math