3 ms·

Has AI advanced from the 1990s? Or has there been a change in the environment it can be applied to, with the rise of big data, big infrastructure - big everythi

by aryehof 10y ago

Has AI advanced from the 1990s? Or has there been a change in the environment it can be applied to, with the rise of big data, big infrastructure - big everything?

thedevil 10y ago
They learned how to train deep nets ten years ago (and are getting better fast as new techniques are invented). As a result, neural nets started taking over. Big Everything is important, but without the new techniques for deep nets, AI would not be so sexy right now.
- edblarney 10y ago
  This should not be downvoted, he is essentially correct. 'AI' has always been around, but the 'big leap' we've seen recently is entirely due to researchers finally able to make Neural Nets actually work :). 'Deep Learning' really refers to a specific kind of Neural Network. The first application was image recognition. It's also used a lot now in voice recognition - and they're trying to jam it into everything possible. I think it's basically fair to say that at least in 2016 AI pretty much boils down to Deep Learning / Neural Nets to move something more classical along in terms of capability. What 'AI' means changes over time, but I think it's safe to say the renewal and hype is based on the 'big eureka' in Neural Nets and their relevant applications.
  - _delirium 10y ago
    Even that I don't think involved much of a "eureka" moment, but mostly incremental improvements due to the environment changing. A lot of neural-net techniques from the '80s and '90s that didn't work that well at the time, now work well because we have sufficient data and computation power (especially via GPUs) to make them work. There have been improvements as well of course, but they're evolutionary improvements, while the big breakthrough was big data + big compute.
tossaway322 10y ago
aryehof asks>"Has AI advanced from the 1990s?" No. But lots of progress in pattern recognition & classification. So whatever AI exists can see better, hear better, speak better. Unfortunately nothing remotely approaching the intelligence of a cat or dog, much less a human being, has appeared except in very limited domains.
- AlexCoventry 10y ago
  nothing remotely approaching the intelligence of a cat or dog, much less a human being, has appeared except in very limited domains. Even machines with the intelligence of a fruit fly or lizard are useful, if you can industrialize the application of them. That is essentially how Alpha Go works.
Dzugaru 10y ago
Computers are far better now in solving complex tasks widely considered as requiring intelligence like video captioning and speech recognition, automatic driving and playing complex games. The change is due to hardware engineering that enabled enormous computing and data acquision/storage power compared to 1990s and a possibility to conduct machine intelligence experiments at unprecedented scale and speed which lead to better algorithms. If, instead, you want to know when or if AI will match human intelligence or become self-consciuos - sadly no one knows that. We may be "advancing" in totally wrong direction.

by aryehof 10y ago

Has AI advanced from the 1990s? Or has there been a change in the environment it can be applied to, with the rise of big data, big infrastructure - big everything?

thedevil 10y ago

They learned how to train deep nets ten years ago (and are getting better fast as new techniques are invented). As a result, neural nets started taking over. Big Everything is important, but without the new techniques for deep nets, AI would not be so sexy right now.

edblarney 10y ago
This should not be downvoted, he is essentially correct. 'AI' has always been around, but the 'big leap' we've seen recently is entirely due to researchers finally able to make Neural Nets actually work :). 'Deep Learning' really refers to a specific kind of Neural Network. The first application was image recognition. It's also used a lot now in voice recognition - and they're trying to jam it into everything possible. I think it's basically fair to say that at least in 2016 AI pretty much boils down to Deep Learning / Neural Nets to move something more classical along in terms of capability. What 'AI' means changes over time, but I think it's safe to say the renewal and hype is based on the 'big eureka' in Neural Nets and their relevant applications.
- _delirium 10y ago
  Even that I don't think involved much of a "eureka" moment, but mostly incremental improvements due to the environment changing. A lot of neural-net techniques from the '80s and '90s that didn't work that well at the time, now work well because we have sufficient data and computation power (especially via GPUs) to make them work. There have been improvements as well of course, but they're evolutionary improvements, while the big breakthrough was big data + big compute.

tossaway322 10y ago

aryehof asks>"Has AI advanced from the 1990s?" No. But lots of progress in pattern recognition & classification. So whatever AI exists can see better, hear better, speak better. Unfortunately nothing remotely approaching the intelligence of a cat or dog, much less a human being, has appeared except in very limited domains.

AlexCoventry 10y ago
nothing remotely approaching the intelligence of a cat or dog, much less a human being, has appeared except in very limited domains. Even machines with the intelligence of a fruit fly or lizard are useful, if you can industrialize the application of them. That is essentially how Alpha Go works.

Dzugaru 10y ago

Computers are far better now in solving complex tasks widely considered as requiring intelligence like video captioning and speech recognition, automatic driving and playing complex games. The change is due to hardware engineering that enabled enormous computing and data acquision/storage power compared to 1990s and a possibility to conduct machine intelligence experiments at unprecedented scale and speed which lead to better algorithms. If, instead, you want to know when or if AI will match human intelligence or become self-consciuos - sadly no one knows that. We may be "advancing" in totally wrong direction.