I wouldn’t want a very technical book,just something which I can read in my free time.Also, I have read On Intelligence
Here are a few of mine that come to mind.
Kurzweil - How to Create a Mind - https://www.amazon.com/How-Create-Mind-Thought-Revealed/dp/1491518839 - I read this around the same time as On Intelligence, it reinforced a few of the same ideas.
Clark - Surfing Uncertainty - https://www.amazon.com/Surfing-Uncertainty-Prediction-Action-Embodied/dp/0190217014 - A book about prediction, predictive coding, and other related concepts in biological and artificial learning systems.
Eagleman - Incognito: The Secret Lives of the Brain - https://www.amazon.com/Incognito-Secret-Lives-David-Eagleman/dp/0307389928 - A book more about the brain than AI, but contains many excellent anecdotes and results about the subconscious and aspects of thinking that we don’t often notice or consider.
Kahneman - Thinking, Fast and Slow - https://www.amazon.com/Thinking-Fast-Slow-Daniel-Kahneman/dp/0374533555 - Again, not about AI, but very relevant to brains and the kinds of processes (fast, intuitive; slow, deliberative) that are responsible for how we work.
Hofstadter - Gödel, Escher, Bach: An Eternal Golden Braid - https://www.amazon.com/Gödel-Escher-Bach-Eternal-Golden/dp/0465026567 - Not actionable information about the brain or AI really, but this was a pretty revolutionary book for me in terms of thinking about nested and hierarchical structures in art, mathematics, and the world.
Hauser - Wild Minds: What Animals Really Think - https://www.amazon.com/Wild-Minds-Animals-Really-Think/dp/080505670X - This one comes with a caveat: Marc Hauser was investigated for academic misconduct and lost his job for it. But most of the work in this book is by other reputable scientists, and the book is a great discussion about how much we underestimate the actual cognitive capabilities of nonhuman animals.
Pfeifer & Bongard - How the Body Shapes the Way We Think: A New View of Intelligence - https://www.amazon.com/How-Body-Shapes-Way-Think/dp/0262162393 - On the embodied view of cognition and the importance of the morphology of the body for intelligent behavior. As a roboticist this has been very influential on my thinking.
Thought I’d reply to this for my first post, since I love to study AI books.
IMHO, metaphor plays a big role wrt intelligence and ultimately AI. Just a few non-technical books not to miss. Of course, just because the equations are few doesn’t mean that the ideas are small.
Hofstadter - Surfaces and Essences, I am a Strange Loop
In my opinion, at least, metaphor is an important component of thought, and DH brings it home with these two, especially his latest, S&E.
Geary - I is An Other
In the image of Lakoff’s work. I haven’t actually completed Lakoff’s oeuvre, but I did complete this book which gives a summary. Lakoff would be another great author to tackle if you’re into analogical AI methods.
Steven Pinker - How the Mind Works, The Language Instinct
Technically not AI books, but certainly about the brain and language, and how the brain creates language. The “instinct” Pinker writes about can be plausibly implemented with a hierarchical structure. So that’s interesting.
Domingos - The Master Algorithm
An excellent recent summary of all sub-fields of AI.
Pribram - The Form Within
A little known book about consciousness and the history of neuroscience. The late Pribram believed in a holographic consciousness. Well, what is SDR other than a holographic representation? A hologram distributes memory across the whole storage space. So does an SDR.
McGilchrist - The Master and His Emissary
Another cog-sci book, not so much AI, but an excellent book.
Marcus - The Future of the Brain
A recent summary of where neuroscience / AI research is.
A nice little book I have read recently.
Common Sense, the Turing Test, and the Quest for Real AI – February 24, 2017
by Hector J. Levesque - https://www.amazon.com/Common-Sense-Turing-Quest-Press/dp/0262036045 - Very recent book, short and packed with some non technical yet powerful thoughts about AI, its promises and its limits.
And something in the “oposite side”? Something in the line of
Penrose, The Emperor’s New Mind… (but more recent) ?
When the book you mention, “How the Mind Works,” came out in 1997, I looked into it and could not find any real explanation of how the brain creates a mind. I was already quite far along in my work as an independent scholar in artificial intelligence, and I had created an animation of my theory – http://ai.neocities.org/theory.html – of how the mind works – which I have now ujploaded as part of this comment. At http://strongai.quora.com/AI-Mind-Maintainer I propose the career-field of AI Mind Maintainer. My own work is somewhat similar to Hierarchical Temporal Memory (HTM) and I have created six AI Mind programs in the course of my work. -Arthur
I just found this book which explains the error correcting effect I saw when experimenting with associative memory.
What I found though is if you keep reapplying the correction the recall drifts away to a different attractor state. Rather inconvenient. There probably are some ways around that. I didn’t make enough of an effort.
Kurzweil is a self proclaimed Futurist ,he doesnt know anything about Neuroscience or Biological approach of Machine based intelligence but i have to agree that he has few interesting opinions about future.
I would expect some cool stuff from this book , if anybody wants to check it out. I am not sure how much of htm / spatial pooling / temporal pooling of the stuff these guys actually use , but looks like the professor is onto something.
I am reading this right now and it’s fascinating. I think it fits nicely with HTM theory as far as mappings and representations are spoken of.
“Neuromancer” by William Gibson. Don’t underestimate the significance of
Do nuclear weapons have anything to do with AI? I would say yes because they are certainly a motivation to get to sentient AI before hubris and ignorance embracing decadence create a hell on earth. I found a very readable historical book:
I hate point this out but mixing AI and nukes seems to be the lead-in to numerous post nuclear dystopian stories.
Like the Forbin Project for example. Anyway you have a species programmed for antiquity, full of emotions and rages with access to nukes.
I don’t need a computer saying that if I kill more of them then them of us it’s a win so start the war.
Or some other bad math or programming bug.
And there’s always the old standby: it’s in diagnostic mode and the safeguards are off.
So many paths to a post nuclear dystopian future.
I’m gonna answer myself here
Anything coming from Dreyfus is mind-blowing.
Anyway, I think that their ideas are not necessarily conflicting with HTM (learn, detect, predict). Perhaps the latest world-modeling-approach from Numenta is.
This one does not explicitly relate itself to Artificial Intelligence, however, the author’s theories about the mind are realistic (by introspection) and can even be realized using computer simulation. For example, a simulated memory machine. This is a brilliant book for non-Neuro people who are extremely interested in how the mind works. The author’s analogies are excellent.
P.S. This is an awesome topic, thanks for asking @Niszoig, now I’ve got more books to read.
While not AI per se, more on the biological side, I think the book Principles of Neural Design was a light read and relevant.