It took a while but i eventually made it through to the end of ‘Shadows of the Mind‘. Most thought provoking book i’ve read in a long time.
As you can see from that link it’s rather controversial. The AI community has spent a great deal of time belittling the idea that it is impossible to build a computational (turing machine based) consciousness. Penrose, however, is more than up to the task of defending his work. It’s a rather strange situation for me, a little like the case of nuclear fusion – the big breakthrough in AI is always just around the corner.
Someone coming at this believing the consensus point of view (of AI) would be surprised that the massive increase in computing power over the last 20 years hasn’t resulted in a corresponding increase in AI ability. Maybe the comparison isn’t fair, but surely by this point we’d be able to have a computer with general inquisitiveness (for example) of a new born. Not even close.
All of which makes it all the more surprising to me that a book that attempts to plot an alternative way forward is so badly received. From my perspective, outside of the (dismal) science of computing (honestly, it’s as much a “science” as economics!) it seem pretty logical. If there are areas of human thought that (provably) cannot be reduced down to algorithmic operations, isn’t it time to consider that maybe trying to reduce all human thought down to algorithmic operations isn’t the answer? Reading some of the criticism (esp. McDermott, linked above) you’d have to wonder if he’s actually read the book, or is simply offended by the idea that there may be limitations to what hard AI can achieve!
For my part, i found the mathematical formalism stuff difficult to follow (but well enough explained that i could get by). The section on quantum effects, superposition, reduction was something of a flashback to university lectures, and therefore quite good fun to relive. And the final section on the micro (or is it nano?) structure of the brain amazingly thought provoking.
I don’t know if Penrose’s specific conclusions (that quantum effects might play a role in the process of consciousness) are all that controversial. There are obviously holes in our understanding of brain function. Therefore putting forward a hypothesis that might fill in some of those holes, even if it is at odds with the existing thinking, is just what is needed. This isn’t like climate denial where most of the answers are known, but unfortunately unpalatable to many who would resist the logical conclusions. The brain really is deeply mysterious, and therefore new ideas should be welcomed in usual scientific spirit!