Is Watson the most advanced super computer in existence?

[DarthTom]

I was watching the faux documentary, "alien planet," that details the fictional exploration of an alien planet in the future by using AI like robots and intelligence. [very good BTW if you haven't seen it]

Alien Planet

And it got me thinking is Watson the human beating super computer on Jeporady even close to such AI, and could hypothetically Watson by itself pilot the exploration of a probe to even one of our planets in this solar system with little or no human intervention?

Secondly, is Watson the most advanced quasi AI we have currently? Or is there better out there?

 

[Lindley]

As I understand it, Watson is just an expert system. Nothing too revolutionary there. Someone correct me if I'm wrong, of course.

 

[JarodRussell]

Watson made logical connections to answer vocal input questions. That doesn't enable it to pilot a spacecraft. I don't think Watson could even play chess because it's just not its purpose. In turn, Deep Blue, the computer that beat Kasparow, can't answer trivia questions. It is one AI that excells in a specific field.

 

[DarthTom]

Watson made logical connections to answer vocal input questions. That doesn't enable it to pilot a spacecraft. I don't think Watson could even play chess because it's just not its purpose. In turn, Deep Blue, the computer that beat Kasparow, can't answer trivia questions. It is one AI that excells in a specific field.

With the programming to pilot a space craft it could. In fact - with the programming to do virtually anything it could - correct?

The larger question is for me:

1. Can Watson - and systems like it - 'learn,' and make decisions based on information?

2. Is Watson the most advanced computer system in existence in terms of its ability to both interact with humans and in terms of computational speeds?

 

[JarodRussell]

Well, yes, with the right programming, any computer can do anything.

Watson is not a magical machine that can read a flight manual and suddenly can fly a plane. It's a first step towards a good search engine, and that's all. It's not even able to properly interact with humans, because that's not what it's designed for. It takes questions, and searches for answers in a big database, and that makes it an excellent contestant in Jeopardy. But not in a conversation.

 

[DarthTom]

Watson is not a magical machine that can read a flight manual and suddenly can fly a plane. It's a first step towards a good search engine, and that's all.

I'm not trying to say it's a magical machine. But I'd hope in terms of its ability to be a 'search engine,' is slightly above [well about 1 million times ahead of ] internet explorers connection to google.

Maybe I'm miss stating the question. Can Watson learn from its mistakes?

edited to add:

It's kind of a shame they are now having Watson act as a customer service agent

 

[Robert Maxwell]

Watson made logical connections to answer vocal input questions. That doesn't enable it to pilot a spacecraft. I don't think Watson could even play chess because it's just not its purpose. In turn, Deep Blue, the computer that beat Kasparow, can't answer trivia questions. It is one AI that excells in a specific field.

With the programming to pilot a space craft it could. In fact - with the programming to do virtually anything it could - correct?

The larger question is for me:

1. Can Watson - and systems like it - 'learn,' and make decisions based on information?

2. Is Watson the most advanced computer system in existence in terms of its ability to both interact with humans and in terms of computational speeds?

Flying spaceships is actually pretty easy for computers, because it involves fairly straightforward physics calculations and a limited number of decisions to make (e.g. which engines to fire at which time to reach point B.) A spacecraft autopilot is quite a bit less complicated than a system like Watson, so it's a bad example.

1. Expert systems can learn more information based on the rules and associations they already have. Depending on how fuzzy its logic is, it could have relations like the following:

* All mammals have hair.
* All mammals are animals.
* All dogs are mammals.
* Yorkshire terrier is a breed of dog.
* A breed is a subcategory/type.

Conclusions:

* All Yorkshire terriers are dogs, mammals, animals, and have hair.

That's a very simple example, but it's relations like these that make up the knowledge of a system like Watson.

2. That's a poor question because computers interact with humans in all kinds of ways. If you mean carrying on a conversation with humans, well, programs like ELIZA remain perennially popular despite not being all that advanced. They are just human enough to fake it. Alice is a more recent example that's good at conversing with humans, albeit in text form. Speech recognition is pretty good these days, so it wouldn't be hard to make a talkie version.

Computational speed is less relevant than memory access strategies. Watson's uniqueness is in the vast amount of information it can sort through quickly, not how fast its CPU is. It is far more data-bound. So, that means very fast (and ample) memory, with fast CPUs being of secondary importance for its purposes.

 

[Pavonis]

Can Watson learn from its mistakes?

Maybe. Depends on what you mean by "learn". I doubt it can, though. Watson is not artificially intelligent in the sense that science fiction has lead to expect. It's at best merely pseudo-intelligent, in that it does a good job of looking like it's intelligent, but it's just a really fast dumb computer with a lot of data behind it. I don't think it can learn like a human.

 

[DarthTom]

Flying spaceships is actually pretty easy for computers, because it involves fairly straightforward physics calculations and a limited number of decisions to make (e.g. which engines to fire at which time to reach point B.) A spacecraft autopilot is quite a bit less complicated than a system like Watson, so it's a bad example.

1. Expert systems can learn more information based on the rules and associations they already have. Depending on how fuzzy its logic is, it could have relations like the following:

* All mammals have hair.
* All mammals are animals.
* All dogs are mammals.
* Yorkshire terrier is a breed of dog.
* A breed is a subcategory/type.

Conclusions:

* All Yorkshire terriers are dogs, mammals, animals, and have hair.

That's a very simple example, but it's relations like these that make up the knowledge of a system like Watson.

2. That's a poor question because computers interact with humans in all kinds of ways. If you mean carrying on a conversation with humans, well, programs like ELIZA remain perennially popular despite not being all that advanced. They are just human enough to fake it. Alice is a more recent example that's good at conversing with humans, albeit in text form. Speech recognition is pretty good these days, so it wouldn't be hard to make a talkie version.

Computational speed is less relevant than memory access strategies. Watson's uniqueness is in the vast amount of information it can sort through quickly, not how fast its CPU is. It is far more data-bound. So, that means very fast (and ample) memory, with fast CPUs being of secondary importance for its purposes.

Thanks Robert Maxwell for the information. In the docu-fiction Alien Planet the probes investigating the planet are claimed to have the 'intelligence,' of a 4 year old. Does Watson and systems like it possess that level of intelligence?

BTW, I'm sure you've read about it. Japan's talking space robot that recently joined the ISS, Kirobo, is another interesting entry in additon to Watson into human - computer interaction.

And could a system like Watson - or even Watson itself - with enough programming successfully investigate a planet with little or no human intervention?

In otherwords - do we possess the technology today to send a probe to an alien planet that would run mostly autonomous of human intervention?

 

[Robert Maxwell]

Indeed. It doesn't "learn" in the sense humans do. It has to be told it is wrong, and in what way, then programmed not to be wrong next time.

I'm not aware of it having any ability to be told it's wrong and then figuring out why and correcting itself. I remember reading that, for any question it has to answer, it builds a list of likely answers, arranged by a confidence number. If it's told one of them is wrong, the confidence number of that answer would become 0 and it would never use that answer again. But the other answers could also all be wrong!

Flying spaceships is actually pretty easy for computers, because it involves fairly straightforward physics calculations and a limited number of decisions to make (e.g. which engines to fire at which time to reach point B.) A spacecraft autopilot is quite a bit less complicated than a system like Watson, so it's a bad example.

1. Expert systems can learn more information based on the rules and associations they already have. Depending on how fuzzy its logic is, it could have relations like the following:

* All mammals have hair.
* All mammals are animals.
* All dogs are mammals.
* Yorkshire terrier is a breed of dog.
* A breed is a subcategory/type.

Conclusions:

* All Yorkshire terriers are dogs, mammals, animals, and have hair.

That's a very simple example, but it's relations like these that make up the knowledge of a system like Watson.

2. That's a poor question because computers interact with humans in all kinds of ways. If you mean carrying on a conversation with humans, well, programs like ELIZA remain perennially popular despite not being all that advanced. They are just human enough to fake it. Alice is a more recent example that's good at conversing with humans, albeit in text form. Speech recognition is pretty good these days, so it wouldn't be hard to make a talkie version.

Computational speed is less relevant than memory access strategies. Watson's uniqueness is in the vast amount of information it can sort through quickly, not how fast its CPU is. It is far more data-bound. So, that means very fast (and ample) memory, with fast CPUs being of secondary importance for its purposes.

Thanks Robert Maxwell for the information. In the docu-fiction Alien Planet the probes investigating the planet are claimed to have the 'intelligence,' of a 4 year old. Does Watson and systems like it possess that level of intelligence?

And could a system like Watson - or even Watson itself - with enough programming successfully investigate a planet with little or no human intervention?

In otherwords - do we possess the technology today to send a probe to an alien planet that would run mostly autonomous of human intervention?

Today, NASA sends semi-autonomous probes to other planets. The Mars rovers cannot be controlled in real-time, for instance. Instead, NASA sends them instructions to be executed--like "go to this location and sample the soil"--and the rover figures out how to get there and deals with any obstacles along the way. It makes its own decisions about how to best achieve its objectives, given its equipment and circumstances.

But you have to realize that this does not constitute "intelligence," in the sense of having a will and being able to make totally independent decisions based on accumulated experiential knowledge.

 

[DarthTom]

But you have to realize that this does not constitute "intelligence," in the sense of having a will and being able to make totally independent decisions based on accumulated experiential knowledge.

Thanks. So in short we do have the level of 'AI,' [for lack of a better term] that can self-direct and as described in the docu-fiction Alien Planet like a 4 year old must be told what's harmful and what's not but can to some degree 'learn that,' through trial and error?

Unrelated question to the TV show Alien Planet:

Is Watson as sophisticated in your opinion as HAL was as depicted in 2001?

In the show - if you haven't seen it - one of the probes is programmed to take greater risks versus the other probe by the super computer oribiting the planet performing the analysis. And they were programmed with specific objectives e.g. one probe to search for plant life and the other animal life.

I suppose that also could be programmed into our own probes on Mars - to be more cautious versus less in performing autonomous executions?

 

[Robert Maxwell]

But you have to realize that this does not constitute "intelligence," in the sense of having a will and being able to make totally independent decisions based on accumulated experiential knowledge.

Thanks. So in short we do have the level of 'AI,' [for lack of a better term] that can self-direct and as described in the docu-fiction Alien Planet like a 4 year old must be told what's harmful and what's not but can to some degree 'learn that,' through trial and error?

Unrelated question to the TV show Alien Planet:

Is Watson as sophisticated in your opinion as HAL was as depicted in 2001?

In the show - if you haven't seen it - one of the probes is programmed to take greater risks versus the other probe by the super computer oribiting the planet performing the analysis. And they were programmed with specific objectives e.g. one probe to search for plant life and the other animal life.

I suppose that also could be programmed into our own probes on Mars - to be more cautious versus less in performing autonomous executions?

HAL is more advanced, in AI terms, because it is sentient and has a desire to preserve itself. Watson doesn't even have the sentience of a cockroach. If you went to unplug it, it would do... well, nothing.

It's not difficult to program probes such that some of them will take greater risks than others. Again, these are just decision trees/matrices, which are well-established in computer science and pretty straightforward to quantify.

 

[DarthTom]

HAL is more advanced, in AI terms, because it is sentient and has a desire to preserve itself. Watson doesn't even have the sentience of a cockroach. If you went to unplug it, it would do... well, nothing.

Doesn't the film 2010 attempt to clean the self preservation issue up by Dr. Chandra essentially explaining away HAL's behavior as a 'h-mobious loop,' whereby the computer became 'paranoid' because it was programmed to keep a secret which caused the glitch?

In other words, if programmers programmed Watson with a secret that they instructed it to keep secret at any an all costs would Watson make similar errors as HAL does in 2001 without being classified as sentient?

 

[Robert Maxwell]

HAL is more advanced, in AI terms, because it is sentient and has a desire to preserve itself. Watson doesn't even have the sentience of a cockroach. If you went to unplug it, it would do... well, nothing.

Doesn't the film 2010 attempt to clean the self preservation issue up by Dr. Chandra essentially explaining away HAL's behavior as a 'h-mobious loop,' whereby the computer became 'paranoid' because it was programmed to keep a secret which caused the glitch?

The Hofstadter-Moebius loop is fiction, yo. A real computer doesn't become "paranoid." One of two things will happen in a computer with conflicting objectives: one of them will take precedence by design or accident, or the system will be stuck in an infinite loop, powerless to act. (The latter would probably lead to a system failure and require a reset.)

 

[Pavonis]

Thanks. So in short we do have the level of 'AI,' [for lack of a better term] that can self-direct and as described in the docu-fiction Alien Planet like a 4 year old must be told what's harmful and what's not but can to some degree 'learn that,' through trial and error?

I think you missed a "not" in there, or else you have completely misunderstood. We do not have anything remotely like the probes in Alien Planet. There's a reason the story was set in the 22nd or 23rd centuries - it's science fiction.

In other words, if programmers programmed Watson with a secret that they instructed it to keep secret at any an all costs would Watson make similar errors as HAL does in 2001 without being classified as sentient?

That's a weird question. Looking to commit homicide with a computer?

 

[DarthTom]

The Hofstadter-Moebius loop is fiction, yo.

I gottcha. I'm not a computer expert - you are - so, the question.

A real computer doesn't become "paranoid." One of two things will happen in a computer with conflicting objectives: one of them will take precedence by design or accident, or the system will be stuck in an infinite loop, powerless to act. (The latter would probably lead to a system failure and require a reset.)

I ha - CTL - ALT - DEL.

I suppose then in an autonomous enviornment conflicting objectives would become a problem to program for with a lot of unknowns?

 

[DarthTom]

I think you missed a "not" in there, or else you have completely misunderstood. We do not have anything remotely like the probes in Alien Planet. There's a reason the story was set in the 22nd or 23rd centuries - it's science fiction.

Of course. The premise of the show is to build upon our current level of technology [AI] and project forward as a docu-fiction.

I was just wondering to what level as depicted on the show ours has reached the one's as seen on the show. And I was wondering generally if Watson is the most advanced super computer on the planet today and if it is even close to that level within 80 to 100 years of advancement.

 

[Pavonis]

No. Watson is not the "most advanced" super computer on the planet. It's just a really good search engine, with good publicity.

 

[DarthTom]

No. Watson is not the "most advanced" super computer on the planet. It's just a really good search engine, with good publicity.

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?

 

[Robert Maxwell]

Watson, not having an ego, doesn't need to have it stroked.