2
$\begingroup$

I took a few AI courses in college (1999-2003), and we used the first edition of AI: A Modern Approach. We covered a lot of topics and programming, including classical AI, neural networks, and temporal difference learning.

Over the past few years, AI has had a resurgence.

Is this resurgence due to new AI theory or just better (more) computational power and data so that the theories (e.g., neural networks) to be more effective? Or is it a combination of both?

I want to get up to speed on what's happened in AI since the early 2000s, and I want to know what to cover -- what is the most significant advancement?

$\endgroup$
2
$\begingroup$

To vastly oversimplify a lot of the progress of modern research in AI/neural nets, the recent advances stems from applicational improvements of the back propagation algorithm.

ex: our recent possession of big data (it turns out that back-prop works very well if you have lots of data to feed into the network), increased computing power to harness that data, and creative network designs to leverage that computing power (deep learning).

So far I haven't found a good source that integrates the knowledge of current practice of neural networks into the classical study of AI nicely. But if you are still interested you can read this modern classic textbook on neural networks and make your own judgements.

http://www.deeplearningbook.org

$\endgroup$
  • $\begingroup$ Thanks. From your response and the book, it seems that our increased computing power and data plus our refinement of machine learning / neural network algorithms is what has put AI back in the spotlight. $\endgroup$ – Brian May 18 '17 at 20:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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