[SATLUG] OT: Linux super computer to compete on Jeopardy

Christopher Connell connell.christopher at yahoo.com
Sat Feb 19 18:13:20 CST 2011

That's a good point Walt I didn't even think that Watson could begin processing the answer as soon as the question is displayed. I also read that Watson, unlike Jennings or Rutter, can consistently press the buzzer at the exact time that the question can be answered (When Alex is finished reading). This article ( http://www.tnr.com/article/83337/ibm-watson-computer-jeopardy ) talks about Watson's likely success on hitting the buzzer faster. 

This link is an interactive game against what the NYTimes says is Watson, but I don't think it really is. 


--- On Sat, 2/19/11, Walt DuBose <dubose at texas.net> wrote:

From: Walt DuBose <dubose at texas.net>
Subject: Re: [SATLUG] OT: Linux super computer to compete on Jeopardy
To: "The San Antonio Linux User's Group Mailing List" <satlug at satlug.org>
Date: Saturday, February 19, 2011, 4:50 PM

On 02/19/2011 05:50 AM, Christopher Connell wrote:
> I watched it when it came on TV and though he answered the same as Jennings (e.g. J: What is the 20's? W: What is the 1920's? Alex: No that's what Ken said. )
> Still can't believe that Watson scored $50,000 more than Rutter and Jennings. Heck combined he still made 31,547 more. Since Watson runs on Linux that really puts other OS's to shame. :-D
> --- On Tue, 2/15/11, Enrique Sanchez<esanchezvela.satlug at gmail.com>  wrote:
> From: Enrique Sanchez<esanchezvela.satlug at gmail.com>
> Subject: Re: [SATLUG] OT: Linux super computer to compete on Jeopardy
> To: "The San Antonio Linux User's Group Mailing List"<satlug at satlug.org>
> Date: Tuesday, February 15, 2011, 11:20 AM
> On Fri, Feb 11, 2011 at 7:45 AM, Borries Demeler
> <demeler at biochem.uthscsa.edu>  wrote:
>>> No the AP video I posted says that watson is that size. Servers that large =
>>> no doubt require a lot of power and something that complex is no doubt expe=
>>> nsive so I just put all that together.=20
>> Check out:
>> https://www.utexas.edu/news/wp-content/uploads/ranger_supercomputer.jpg
>> http://www.utexas.edu/news/images/2008/research/2946_ranger_280_187.jpg
> http://www.youtube.com/watch?v=ZLdkJpAtt1I
> best regards,
> esv.

For one thing the computer has more time...as the question/statement is 
being read, if there are key words in it before the host gives the Ok, 
the computer is already getting the answer...then if the probability is 
goood that the computer has a viable answer, it starts the ring-the-bell 
action before it answers.  This could give the computer maybe 2 seconds 
to finish its answer before having to give the answer.  And it might 
even keep on trying to find the answer in case its original answer was 
wrong and it had a second change at getting the answer right.

Its really no big thing as most of the newer air-to-air missiles use 
time compress to track their target.

SATLUG mailing list
SATLUG at satlug.org
http://alamo.satlug.org/mailman/listinfo/satlug to manage/unsubscribe
Powered by Rackspace (www.rackspace.com)


More information about the SATLUG mailing list