5 ms·

Whenever we figure out how to do something, we stop calling it AI or AGI. If the trend continues, will we eventually have a general AI, but won't consider it an

by Swizec 10y ago

Whenever we figure out how to do something, we stop calling it AI or AGI. If the trend continues, will we eventually have a general AI, but won't consider it anything special? Will it have been just a small incremental step by then?

kowdermeister 10y ago
I think it will need some kind of breakthrough. Current advancements are probably incremental as you stated, but having an AGI might need some new theory we don't have currently.
- xiphias 10y ago
  Deep learning is the opposite of incremental. For a long time it was not clear whether/how we can learn multi layer networks efficiently. ImageNet changed everything.
  - bonoboTP 10y ago
    Machine learning people basically agree that there weren't any big breakthroughs in deep learning. The success and the hype is mostly a combination of more computing power and more data. The algorithms (convolutional neural network etc.) were invented back in the 1980s and even earlier. There have been some improvements but they are incremental indeed. More use of ReLU, dropout etc. But it's not a new paradigm at all.
    - ebalit 10y ago
      They weren't any recent breakthroughts. But LSTM and ConvNet are breakthroughts. It just took a lot of times to prove it.
      - bonoboTP 10y ago
        Convnets follow pretty naturally from multilayer perceptrons. Perhaps backpropagation was a breakthrough, enabling the training of ANNs on data, instead of hand-tuning. But the idea of neural nets is very old, going back to Rosenblatt and connectionism.
bbctol 10y ago
I think deep, natural language processing will be unambiguous: if you create a machine that says "Yes, I am intelligent, thanks for asking" in a way indistinguishable from a human, it would be hard to disagree. On the other hand, it's entirely possible that that goal will take so much longer than others we'll have incredibly strong AIs affecting our lives before we notice.
extesy 10y ago
Yes, this is called AI effect: https://en.wikipedia.org/wiki/AI_effect https://en.wikipedia.org/wiki/AI_effect

by Swizec 10y ago

kowdermeister 10y ago

I think it will need some kind of breakthrough. Current advancements are probably incremental as you stated, but having an AGI might need some new theory we don't have currently.

xiphias 10y ago
Deep learning is the opposite of incremental. For a long time it was not clear whether/how we can learn multi layer networks efficiently. ImageNet changed everything.
- bonoboTP 10y ago
  Machine learning people basically agree that there weren't any big breakthroughs in deep learning. The success and the hype is mostly a combination of more computing power and more data. The algorithms (convolutional neural network etc.) were invented back in the 1980s and even earlier. There have been some improvements but they are incremental indeed. More use of ReLU, dropout etc. But it's not a new paradigm at all.
  - ebalit 10y ago
    They weren't any recent breakthroughts. But LSTM and ConvNet are breakthroughts. It just took a lot of times to prove it.
    - bonoboTP 10y ago
      Convnets follow pretty naturally from multilayer perceptrons. Perhaps backpropagation was a breakthrough, enabling the training of ANNs on data, instead of hand-tuning. But the idea of neural nets is very old, going back to Rosenblatt and connectionism.

bbctol 10y ago

I think deep, natural language processing will be unambiguous: if you create a machine that says "Yes, I am intelligent, thanks for asking" in a way indistinguishable from a human, it would be hard to disagree. On the other hand, it's entirely possible that that goal will take so much longer than others we'll have incredibly strong AIs affecting our lives before we notice.

extesy 10y ago

Yes, this is called AI effect: https://en.wikipedia.org/wiki/AI_effect https://en.wikipedia.org/wiki/AI_effect