khiner

My main takeaway: “How can deep learning be good if it’s not symbolic reasoning, you know, like scientists have proven brains work like? We think in symbols, symbols have inherent meaning, why aren’t deep learning learning symbols?! I’ve been saying exactly this for 20 years now! You know, all those deep learning hot shots think they’re so cool, but have they anticipated symbols like I have? Symbols won’t just allow us to play Atari games from raw pixel inputs, they will give us general AI. Somebody please write a paper showing how a symbol based algorithm of some kind is generally intelligent, and also that any connectionist-leaning mechanism is orthogonal to complex adaptive agent behaviors!”

Thanks, glad you enjoyed it! Lately I’ve had a break from full time work, so I could dedicate all my time to them. This one took a little over three weeks I think, which is pretty typical. His next book, Physical Audio Signal Processig, is a pretty huge tomb of a textbook so I imagine it would take much longer even with full time dedication. I started it but have to put it down for a bit to look for work. Even with this Filters book it was easy to get stuck on a single problem for most of the day.

Yes, gotcha. My whole site is a react app with a lot of MathJax formatting and some posts center around JavaScript animations. But mainly the holdup is that it uses async-loaded code-splitting techniques to keep initial page loads down. It would be a big challenge to make it friendly without a lot of dynamic language help. Apologies. (Although I will think about how to make text-only content available as I agree with the idea that JS-only is unnecessarily restrictive when most of the actual content is text and images.)

Changed “remove” to “attenuate”. Thanks again.

Cool! I could see reading this one first for sure, although DFT knowledge definitely helps.

Same here really, a software engineer by trade with a growing interest in audio. I’ve been really impressed with the quality of these books so far. Very meticulous attention to detail with _very_ few errors compared with most other books of its kind.

Are you refering to my site or JOS’s site? Can you elaborate please?

Thanks! I have a similar post for Mathemeatics of the DFT, and I’ve started Physical Audio Signal Processing, but currently distracted by other software projects. I love the books, can’t wait to get back to them.

That is a really good point. “Remove” is a pretty flippant word to use in this context. I’ll update, thanks :)

For a much better and more constructive read toward the effort of a stringent definition of intelligence, I would recommend anything from Shane Legg - most pragmattically this: https://arxiv.org/pdf/1109.5951v2.pdf

I didn't read this at first because I thought it just sounded like it would be some opinionated clickbait full of strawmen and superlatives. But then I caved in and read it and it turns out that my instincts were right. Then I saw the author was Kevin Kelly and I felt sad. Almost as sad as when Steven Hawking said we should discontinue the SETI program because it is most likely that aliens will want to harvest our precious carbon and organs and enslave us if they found out we were here. UNSUBSCRIBE!

I've been working on this app, too! For Android. We should all have pow-wow about our approaches. karl.hiner@gmail.com