What are the current limits of artificial intelligence? & their weaknesses

[Zzzz1]

I added an example in my previous post.

Of course, but I feel like we are arguing about apples and oranges. Your argument is that automation can do many tasks better then 90% of human beings - I'd say that (a) that number is too low and (b) it's not a very difficult benchmark to beat. The majority of human beings pretty much suck at everything, from menial tasks to high cognition activities. It's only natural that you can create an automation that can replace humans in some of those tasks by designing pin-point SML tools.

PS. For the record, I am an investor in a several SML/AI related startup companies so my opinions might be biased

 

[sle]

Hmm, I am not sure to be honest. It is a terminology issue, AI for me is a collective description of any learning process that is automated. SML is admittedly lower in the food chain with regressions, linear classification, minimax, clustering, and the like. So, I would consider it part of AI. Not of the sophisticated and complex kind but nonetheless.

When I was doing my PhD, I've always felt that lumping SML with AI is somewhat wrong (while now I am happy to do that now when VC funding is at stake ).

For example, look at machine translation - SML-based solutions have brought us to a point when machine-translation is semi-usable. Yet it's quality is more about the availability of the parallel corpus and less about the actual context cognition.

 

[Zzzz1]

Can't comment on that, I never worked with textual data. It is an entirely different field with its own challenges and complexities.

When I was doing my PhD, I've always felt that lumping SML with AI is somewhat wrong (while now I am happy to do that now when VC funding is at stake ).

For example, look at machine translation - SML-based solutions have brought us to a point when machine-translation is semi-usable. Yet it's quality is more about the availability of the parallel corpus and less about the actual context cognition.

 

[sle]

Can't comment on that, I never worked with textual data. It is an entirely different field with its own challenges and complexities.

It's just an example where statistical learning and cognition learning are miles away from each other. My point is that what defines intelligence is not excellence in a specific activity, but the reasoning ability. We strongly distinguish between bright children and idiot-savants, don't we?

 

[Humpy]

Didn't I read somewhere that a kid with a laser torch could cause complete chaos on the roads with driver less cars ?

 

[Cdntrader]

Weak AI can be useful, but also dangerous when misapplied or overly trusted: https://tech.slashdot.org/story/17/04/15/0132203/tiny-changes-can-cause-an-ai-to-fail

Strong AI can hopefully replace all of humanity,

Did airplanes replace birds?


There isn't even an accepted definition of intelligence.

 

[nooby_mcnoob]

There is no real AI yet, some clever programs which they class as AI but that's all they are and there not AI.

I personally, don't think we'll ever see real AI!

What is most likely to happen is we will create clever programs classed as AI that will begin working together in ways that we did not directly program. Then they will be combined in various ways creating "brains" that are good at X,Y,Z.

Think about how we work, I can't imagine it's that much different.

 

[userque]

Didn't I read somewhere that a kid with a laser torch could cause complete chaos on the roads with driver less cars ?

Works with human-driven cars as well.

 

[userque]

Can't comment on that, I never worked with textual data. It is an entirely different field with its own challenges and complexities.

I must say, you don't give up easily!

As a custom ML algo coder/trader and also as one who has driven many miles, and seen many dumb distracted and undistracted drivers, I'd rather 'human-drive' in a world where all other drivers were final phase AI (or ML or whatever you desire to call it ), than drive in the current scenario.

I also suspect there would be far fewer accidents in a world where only AI drove cars, than the current scenario.

I remember trying to explain the science behind microwave ovens to my Mother way back when. She had a problem with 'irradiated' food.

My motive was that I wanted a microwave in the house to facilitate the quick and easy cooking of hot dogs. But for that desire, I wouldn't have spent so many days and hours championing the fact that 'microwave ovens are safe--it wouldn't have mattered much to me that she choose to believe that they were unsafe."

Good luck!

 

[Turveyd]

I work with AI every single day, you tard. Do you?

BS, if you worked in AI every day, you'd know no computer system as yet has ever achieved intelligence, which is what the I stands for.

They've got better at making computers loop around and repeat not as simple as before tasks millions of times per second, but that is all there doing.