Meh. I don't think you're giving, "him," enough credit. How many computers do you run accross in everyday life that can beat not 1 but 2 or 3 Jeporady champions using natural language interaction?
None. Why would I need or want to interact with one?
Is Watson the most advanced super computer in existence?
- Thread starter DarthTom
- Start date
None. Why would I need or want to interact with one?
Well, in my case - Watson seems a lot more impressive than Siri is at answering questions.
Siri on tough questions, 'let me go to the web and see if I can find that....'
Watson answers questions provided in a very specific format that have known factual answers.
Siri has to deal with people asking all kinds of random crap. The garbage out is going to be high because the garbage in is high.
Siri has to deal with people asking all kinds of random crap. The garbage out is going to be high because the garbage in is high.
That's true. I just tested your statement by asking Siri as series of straight forward questions e.g. 'What year was Enstein Born,' 'What is Enstein famous for,' and she correctly answered the questions - albiet used Wiki as a reference in repeating the correct answer.
Are you saying that Watson and Siri are equal in capabilities?
Last edited:
Watson only knows what's in its databases, and can search it quickly. Siri can take voice input and use it for a keyword search on the web. They're very different, but of the two, Siri is affordable and more useful for the everyday needs of a person. No one really needs a pocket Watson.
That would be profoundly stupid if Watson wasn't programmed to search the internet as well as its own database to find answers to queries.
But yes I agree Watson [at least for now] isn't a viable economic choice for most people. I'm surprised that IBM hasn't tried to develop and market a competitor to Siri with the Watson interface.
Most of the android voice recognition systems don't hold a candle to Siri in my experience.
Nope. In 3G or lower bandwith areas I get the, "I cannot answer any questions right now," response which obviously indicates that it isn't running locally. You'd think the software though wouldn't go out to the web for the time and date though - but it does.
Well, yes, Siri is a pocket Watson. What I meant was that you don't need to have the entire web stored on a single device when you have Siri to search the web. As far as I know, Watson doesn't do web searches, it just has data stored in servers that it accesses. DarthTom seems to be under the impression that Watson is "superior" to Siri, though I can't fathom why.
Watson is constrained by the time limit of Jeopardy questions, so local storage is a must.
Siri can afford to be a bit slower.
Siri can afford to be a bit slower.
Well for starters, I'm guessing that Siri couldn't beat a Jeporday champion.
Last edited:
Siri could beat a Jeopardy champion, if Apple wanted to spend the time and money on the publicity stunt.
Computers are tools, not competitors. The IBM Watson competition was just to raise publicity for their research, not to produce an artificially intelligent being.
Computers are tools, not competitors. The IBM Watson competition was just to raise publicity for their research, not to produce an artificially intelligent being.
Siri isn't processing the vocal requests locally on your device. If it can't access the "mainframe" back at Apple then it can't understand you.
This thread is really kind of weird. 
This thread is really kind of weird.
Why is that? I watched a program - Alien Planet - that discusses the idea that virtually AI will one day explore nearby planets within the, 'goldi-locks zone,' of nearby star systems on its own.
After watching the show it occured to me that we have super computers - like Watson - that have almost the same capabilities as described on the show.
I was curious if Watson and other computers today reach the level of 'intelligence,' necessary today to perform such tasks.
Robert Maxwell said that many of the tasks as described on the show are possible.
Tangentile to the discussion was how 'powerful,' Watson is relative to other supercomputers.
Is this forum not where science and technology is discussed? Do you have anything to contribute to the conversation?
DarthTom, you don't need to mini-mod here.
iguana is probably a bit put off by, to put it nicely, how not very knowledgeable you are about computers and computing technology. You seem easily impressed. Not that it's not impressive, but there seems to be a common tendency to view technology you don't understand as almost magical in nature. You seem very impressed, for instance, by Watson, going so far as to (initially) think it's the most advanced computer in existence. I assume you came to this conclusion due to its ability to beat other Jeopardy contestants, but those who likened it to Deep Blue have it right. Like Deep Blue, there is nothing magical about Watson, it is just a very powerful, highly specialized computer designed to solve one particular task very efficiently. Its existence is more evolutionary than revolutionary. It's not some kind of turning point in AI. It's a notable achievement, but I'm just saying not to make more of it than it is.
We could send automated probes to other solar systems if we wanted to and were willing to spend the money. The reason we don't is that it would take so long and the benefits, if any, would come several generations down the line. That makes it politically uninteresting, since no one in office (or even alive) today would be able to take credit for it. There's also no good economic incentive.
At this point, as our telescopes get better and better, we'll learn a lot more about extrasolar planets that way, and if we find a really, really promising one, maybe we'll send some probes, but I don't think you appreciate just how long-term a project that is. To put it in perspective, the fastest ships we've built to date--Helios-A and Helios-B--have had a maximum velocity of a bit over 150,000 miles per hour. Light travels at almost 670 million miles per hour. That means the fastest ship we've ever made has gone 0.02% the speed of light. It would take almost 43 years for such a ship to travel one light year. That's over 180 years just to the nearest star.
To see results in any reasonable amount of time, you'd need ships at least 10 times faster, and you'd need to be sure they could survive such high velocities. The computer systems are really the least of our worries with something like that. Maintaining power and velocity through the interstellar medium would be of much greater concern.
Basically: computing power is not what is holding us back from exploring space, and it never really has been. Systems like Watson and Siri are impressive but represent iterative improvements and integration of technologies and concepts that have been around for decades. There is no magic here, just a bunch of real-world reasons why we have achieved some things and not others.
iguana is probably a bit put off by, to put it nicely, how not very knowledgeable you are about computers and computing technology. You seem easily impressed. Not that it's not impressive, but there seems to be a common tendency to view technology you don't understand as almost magical in nature. You seem very impressed, for instance, by Watson, going so far as to (initially) think it's the most advanced computer in existence. I assume you came to this conclusion due to its ability to beat other Jeopardy contestants, but those who likened it to Deep Blue have it right. Like Deep Blue, there is nothing magical about Watson, it is just a very powerful, highly specialized computer designed to solve one particular task very efficiently. Its existence is more evolutionary than revolutionary. It's not some kind of turning point in AI. It's a notable achievement, but I'm just saying not to make more of it than it is.
We could send automated probes to other solar systems if we wanted to and were willing to spend the money. The reason we don't is that it would take so long and the benefits, if any, would come several generations down the line. That makes it politically uninteresting, since no one in office (or even alive) today would be able to take credit for it. There's also no good economic incentive.
At this point, as our telescopes get better and better, we'll learn a lot more about extrasolar planets that way, and if we find a really, really promising one, maybe we'll send some probes, but I don't think you appreciate just how long-term a project that is. To put it in perspective, the fastest ships we've built to date--Helios-A and Helios-B--have had a maximum velocity of a bit over 150,000 miles per hour. Light travels at almost 670 million miles per hour. That means the fastest ship we've ever made has gone 0.02% the speed of light. It would take almost 43 years for such a ship to travel one light year. That's over 180 years just to the nearest star.
To see results in any reasonable amount of time, you'd need ships at least 10 times faster, and you'd need to be sure they could survive such high velocities. The computer systems are really the least of our worries with something like that. Maintaining power and velocity through the interstellar medium would be of much greater concern.
Basically: computing power is not what is holding us back from exploring space, and it never really has been. Systems like Watson and Siri are impressive but represent iterative improvements and integration of technologies and concepts that have been around for decades. There is no magic here, just a bunch of real-world reasons why we have achieved some things and not others.
The other thing is that Watson runs on 750 servers with 4 cores each. I don't know if you could even supply the amount of power needed for a long term mission.
10-4
I don't possess a degree in computer science and my, 'exposure,' to the technology is limited to using Siri, MS Office for business and work an like most people use computers as tools for work.
So yes - sorry - super computing is exciting to me and I don't understand what happens under the proverbial hood of the car so I ask questions.
Any computer that can beat a human being at Jeporday - IMO - is an amazing feat considering that 99% of the human population could not beat the flesh and blood winners either.
edited to add: then again I think it was impressive, funny, and sad when I asked Siri Sunday afternoon the Falcons score she said, "The Falcons got 'crushed,' by Tampa Bay...." - obviously someone in Curpertino feeling th need to add that editorial to the socre. LOL
Last edited:
But Jeopardy is not about intelligence, it's about having vast amounts of trivia knowledge. Pretty much anyone can accumulate that with some time and effort. It's not a question of intellectual ability.
The thing about that is... the Jeopardy task is simply suitable for machines. Just like calculations. Computers beat humans since decades in that area. Because the human brain just isn't designed for that.
On the other hand, it will take a very, very long time until, for example, a computer beats a human in visual recognition. Because that's what the human brain can do with perfection.
