MIT Working on AI Reinvention

Terry Olaes

Terry Olaes

I Used to be the [H] News Guy
Joined
Nov 27, 2006
Messages
4,646
The field of Artificial Intelligence (AI) was founded more than 50 years ago and much of the intervening time seems to have been spent on trying to figure out the bare basics. A project at the Massachusetts Institute of Technology (MIT) is looking to “reboot” this field, leveraging what is already known to make some significant headway. Insert obvious Skynet or Matrix reference here.

The project is “revisiting fundamental assumptions” in all of the areas encompassed by the field of AI, including the nature of the mind and of memory, and how intelligence can be manifested in physical form, says Gershenfeld, professor of media arts and sciences. “Essentially, we want to rewind to 30 years ago and revisit some ideas that had gotten frozen,” he says, adding that the new group hopes to correct “fundamental mistakes” made in AI research over the years.
Click to expand...
 
Would be nice to see some advance AI making its way into commercial video games, having seen any significant improvement in video games AI lately. Still the plain old boring enemy running around
 
It'd be nice if we'd have some AI in freakin stop lights. I mean, why do I have to wait 5 min for a light to turn green at 2am when there is no cross traffic.

I could also see AI being useful in GPS systems for the navy as well as general public. Also, AI would make communication technology a bit better as well.
 
djgizmo said:
It'd be nice if we'd have some AI in freakin stop lights. I mean, why do I have to wait 5 min for a light to turn green at 2am when there is no cross traffic.

I could also see AI being useful in GPS systems for the navy as well as general public. Also, AI would make communication technology a bit better as well.
Click to expand...

Dynamic road management would be a great use of AI.
 
djgizmo said:
It'd be nice if we'd have some AI in freakin stop lights. I mean, why do I have to wait 5 min for a light to turn green at 2am when there is no cross traffic.

I could also see AI being useful in GPS systems for the navy as well as general public. Also, AI would make communication technology a bit better as well.
Click to expand...

That's something that comes down to state-funding. There's been actuators/computers at intersections for quite a while now (I'd say a couple decades at least). If your county/city is too cheap to spend on that though, too bad.

However, the thinking that computers will solve traffic conditions is a bit too optimistic. Remember that humans are still driving cars, and that vehicles and additional O/D's are getting added every day. It doesn't matter how many roads, or how optimized things get b/c more cars are more cars - nothing's gonna solve that.

Things like progression (how far can your car go with constant green lights) - that can be optimized...to an extent. The problem remains that there are simply too many damn cars. Sorry to be an pessimist, but having looked at the numbers, there is never an easy solution.
 
i dont care what they do with AI just as long as it is NOT connected to the internet
 
Mav451 said:
That's something that comes down to state-funding. There's been actuators/computers at intersections for quite a while now (I'd say a couple decades at least). If your county/city is too cheap to spend on that though, too bad.

However, the thinking that computers will solve traffic conditions is a bit too optimistic. Remember that humans are still driving cars, and that vehicles and additional O/D's are getting added every day. It doesn't matter how many roads, or how optimized things get b/c more cars are more cars - nothing's gonna solve that.

Things like progression (how far can your car go with constant green lights) - that can be optimized...to an extent. The problem remains that there are simply too many damn cars. Sorry to be an pessimist, but having looked at the numbers, there is never an easy solution.
Click to expand...

totally missed the point dude. Sitting at a stoplight at 2am with no other cars =/= problems with traffic congestion
 
Ryan711 said:
totally missed the point dude. Sitting at a stoplight at 2am with no other cars =/= problems with traffic congestion
Click to expand...

No, you totally missed the point. That is not a problem that AI is needed to solve. You combine a weight sensor and 20-year old computer technology and you've solved that problem.

Mav451 said:
That's something that comes down to state-funding. There's been actuators/computers at intersections for quite a while now (I'd say a couple decades at least). If your county/city is too cheap to spend on that though, too bad.
Click to expand...
 
Nemesis999 said:
No, you totally missed the point. That is not a problem that AI is needed to solve. You combine a weight sensor and 20-year old computer technology and you've solved that problem.
Click to expand...

He just need to admit that his town sucks :p
 
This "reinvention" of AI is nothing new. They've tried it before. I mean, after AI failed to reach it's goals in the late 60's and early 70's, they kinda gave up. Then they "reinvented" AI with neural nets, and while they do have their uses, neural networks never panned out to be a solve-all. After that, AI just got pigeon-holed into specific applications like data mining and medical diagnosis. This whole "reinventing" thing just sounds like a rehash of past ideas, ideas which never worked, btw.

I think a better solution is to focus more on Artificial General Intelligence (AGI) and focus on how it can be grounded in (relatively) consistent perception and action states (e.g. a body of some sort).
 
You must log in or register to reply here.
Back
Top