1
$\begingroup$

It is common to see articles praising the advances in AI (here is one example).

There is no doubt that significant progress has occurred on all kinds of automation, hard games like chess, etc.

The question is, how much of that progress is due to conceptual proogress, as opposed to increase in processing power? Or, to phrase it in a different way: if today's best AI would be shown to a computer scientist from the 60s, would they be impressed by the algorithms? Or just by the raw power of the hardware?

$\endgroup$
1
  • $\begingroup$ Have you done any research to look for articles? A quick Google search turns up a few. $\endgroup$
    – D.W.
    Commented Mar 24, 2021 at 6:24

1 Answer 1

2
$\begingroup$

AI spans many different areas. If you are actually asking about machine learning, the conventional wisdom is probably that it's a combination of new ideas (e.g., neural networks, architectures, methods for training them), computation power (e.g., GPUs), and data (large labelled datasets).

See, e.g., https://qz.com/1307091/the-inside-story-of-how-ai-got-good-enough-to-dominate-silicon-valley/, https://towardsdatascience.com/the-state-of-ai-in-2020-1f95df336eb0, and probably many more (I didn't spend much time looking).

$\endgroup$
1
  • $\begingroup$ I have read both articles now. The first one simply mentions that an advance was made by improving an algorithm and running it on better hardware. The second one rambles for many paragraphs about algorithms that mean nothing to me, and then near the end briefly mentions that to get real advances a new paradigm is needed; and the author hopes that quantum computing could be it (which I really doubt). None of the articles answers my question really. $\endgroup$ Commented Mar 26, 2021 at 5:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.