Welcome to our discussion of Ray Kurzweil’s How to Create a Mind: The Secret of Human thought Revealed. This book was requested by one my fine readers; I hope you have enjoyed it.
If you aren’t familiar with Ray Kurzweil (you must be new to the internet), he is a computer scientist, inventor, and futurist whose work focuses primarily on artificial intelligence and phrases like “technological singularity.”
Wikipedia really likes him.
The book is part neuroscience, part explanations of how various AI programs work. Kurzweil uses models of how the brain works to enhance his pattern-recognition programs, and evidence from what works in AI programs to build support for theories on how the brain works.
The book delves into questions like “What is consciousness?” and “Could we recognize a sentient machine if we met one?” along with a brief history of computing and AI research.
My core thesis, which I call the Law of Accelerating Returns, (LOAR), is that fundamental measures of of information technology follow predictable and exponential trajectories…
I found this an interesting sequel to Auerswald’s The Code Economy and counterpart to Gazzaniga’s Who’s In Charge? Free Will and the Science of the Brain, which I listened to in audiobook form and therefore cannot quote very easily. Nevertheless, it’s a good book and I recommend it if you want more on brains.
The quintessential example of the law of accelerating returns is the perfectly smooth, doubly exponential growth of the price/performance of computation, which has held steady for 110 years through two world was, the Great Depression, the Cold War, the collapse of the Soviet Union, the reemergence of China, the recent financial crisis, … Some people refer to this phenomenon as “Moore’s law,” but… [this] is just one paradigm among many.
Auerswald claims that the advance of “code” (that is, technologies like writing that allow us to encode information) has, for the past 40,000 years or so, supplemented and enhanced human abilities, making our lives better. Auerswald is not afraid of increasing mechanization and robotification of the economy putting people out of jobs because he believes that computers and humans are good at fundamentally different things. Computers, in fact, were invented to do things we are bad at, like decode encryption, not stuff we’re good at, like eating.
The advent of computers, in his view, lets us concentrate on the things we’re good at, while off-loading the stuff we’re bad at to the machines.
Kurzweil’s view is different. While he agrees that computers were originally invented to do things we’re bad at, he also thinks that the computers of the future will be very different from those of the past, because they will be designed to think like humans.
A computer that can think like a human can compete with a human–and since it isn’t limited in its processing power by pelvic widths, it may well out-compete us.
But Kurzweil does not seem worried:
Ultimately we will create an artificial neocortex that has the full range and flexibility of its human counterpart. …
When we augment our own neocortex with a synthetic version, we won’t have to worry about how much additional neocortex can physically fit into our bodies and brains, as most of it will be in the cloud, like most of the computing we use today. I estimated earlier that we have on the order of 300 million pattern recognizers in our biological neocortex. That’s as much as could b squeezed into our skulls even with the evolutionary innovation of a large forehead and with the neocortex taking about 80 percent of the available space. As soon as we start thinking in the cloud, there will be no natural limits–we will be able to use billions or trillions of pattern recognizers, basically whatever we need. and whatever the law of accelerating returns can provide at each point in time. …
Last but not least, we will be able to back up the digital portion of our intelligence. …
That is kind of what I already do with this blog. The downside is that sometimes you people see my incomplete or incorrect thoughts.
On the squishy side, Kurzweil writes of the biological brain:
The story of human intelligence starts with a universe that is capable of encoding information. This was the enabling factor that allowed evolution to take place. …
The story of evolution unfolds with increasing levels of abstraction. Atoms–especially carbon atoms, which can create rich information structures by linking in four different directions–formed increasingly complex molecules. …
A billion yeas later, a complex molecule called DNA evolved, which could precisely encode lengthy strings of information and generate organisms described by these “programs”. …
The mammalian brain has a distinct aptitude not found in any other class of animal. We are capable of hierarchical thinking, of understanding a structure composed of diverse elements arranged in a pattern, representing that arrangement with a symbol, and then using that symbol as an element in a yet more elaborate configuration. …
I really want to know if squids or octopuses can engage in symbolic thought.
Through an unending recursive process we are capable of building ideas that are ever more complex. … Only Homo sapiens have a knowledge base that itself evolves, grow exponentially, and is passe down from one generation to another.
Kurzweil proposes an experiment to demonstrate something of how our brains encode memories: say the alphabet backwards.
If you’re among the few people who’ve memorized it backwards, try singing “Twinkle Twinkle Little Star” backwards.
It’s much more difficult than doing it forwards.
This suggests that our memories are sequential and in order. They can be accessed in the order they are remembered. We are unable to reverse the sequence of a memory.
Funny how that works.
On the neocortex itself:
A critically important observation about the neocortex is the extraordinary uniformity of its fundamental structure. … In 1957 Mountcastle discovered the columnar organization of the neocortex. … [In 1978] he described the remarkably unvarying organization of the neocortex, hypothesizing that it was composed of a single mechanism that was repeated over and over again, and proposing the cortical column as the basic unit. The difference in the height of certain layers in different region noted above are simply differences in the amount of interconnectivity that the regions are responsible for dealing with. …
extensive experimentation has revealed that there are in fact repeating units within each column. It is my contention that the basic unit is a pattern organizer and that this constitute the fundamental component of the neocortex.
As I read, Kurzweil’s hierarchical models reminded me of Chomsky’s theories of language–both Ray and Noam are both associated with MIT and have probably conversed many times. Kurzweil does get around to discussing Chomsky’s theories and their relationship to his work:
Language is itself highly hierarchical and evolved to take advantage of the hierarchical nature of the neocortex, which in turn reflects the hierarchical nature of reality. The innate ability of human to lean the hierarchical structures in language that Noam Chomsky wrote about reflects the structure of of the neocortex. In a 2002 paper he co-authored, Chomsky cites the attribute of “recursion” as accounting for the unique language faculty of the human species. Recursion, according to Chomsky, is the ability to put together small parts into a larger chunk, and then use that chunk as a part in yet another structure, and to continue this process iteratively In this way we are able to build the elaborate structure of sentences and paragraphs from a limited set of words. Although Chomsky was not explicitly referring here to brain structure, the capability he is describing is exactly what the neocortex does. …
This sounds good to me, but I am under the impression that Chomsky’s linguistic theories are now considered outdated. Perhaps that is only his theory of universal grammar, though. Any linguistics experts care to weigh in?
Within the field of linguistics, McGilvray credits Chomsky with inaugurating the “cognitive revolution“. McGilvray also credits him with establishing the field as a formal, natural science, moving it away from the procedural form of structural linguistics that was dominant during the mid-20th century. As such, some have called him “the father of modern linguistics”.
The basis to Chomsky’s linguistic theory is rooted in biolinguistics, holding that the principles underlying the structure of language are biologically determined in the human mind and hence genetically transmitted. He therefore argues that all humans share the same underlying linguistic structure, irrespective of sociocultural differences. In adopting this position, Chomsky rejects the radical behaviorist psychology of B. F. Skinner which views the mind as a tabula rasa (“blank slate”) and thus treats language as learned behavior. Accordingly, he argues that language is a unique evolutionary development of the human species and is unlike modes of communication used by any other animal species. Chomsky’s nativist, internalist view of language is consistent with the philosophical school of “rationalism“, and is contrasted with the anti-nativist, externalist view of language, which is consistent with the philosophical school of “empiricism“.
Anyway, back to Kuzweil, who has an interesting bit about love:
Science has recently gotten into the act as well, and we are now able to identify the biochemical changes that occur when someone falls in love. Dopamine is released, producing feelings of happiness and delight. Norepinephrin levels soar, which lead to a racing heart and overall feelings of exhilaration. These chemicals, along with phenylethylamine, produce elevation, high energy levels, focused attention, loss of appetite, and a general craving for the object of one’s desire. … serotonin level go down, similar to what happens in obsessive-compulsive disorder….
If these biochemical phenomena sound similar to those of the flight-or-fight syndrome, they are, except that we are running toward something or someone; indeed, a cynic might say toward rather than away form danger. The changes are also fully consistent with those of the early phase of addictive behavior. … Studies of ecstatic religious experiences also show the same physical phenomena, it can be said that the person having such an experiences is falling in love with God or whatever spiritual connection on which they are focused. …
Religious readers care to weigh in?
Consider two related species of voles: the prairie vole and the montane vole. They are pretty much identical, except that the prairie vole has receptors for oxytocin and vasopressin, whereas the montane vole does not. The prairie vole is noted for lifetime monogamous relationships, while the montane vole resorts almost exclusively to one-night stands.
Learning by species:
A mother rat will build a nest for her young even if she has never seen another rat in her lifetime. Similarly, a spider will spin a web, a caterpillar will create her own cocoon, and a beaver will build a damn, even if no contemporary ever showed them how to accomplish these complex tasks. That is not to say that these are not learned behavior. It is just that he animals did not learn them in a single lifetime… The evolution of animal behavior does constitute a learning process, but it is learning by the species, not by the individual and the fruits of this leaning process are encoded in DNA.
I think that’s enough for today; what did you think? Did you enjoy the book? Is Kurzweil on the right track with his pattern recognizers? Are non-biological neocortexes on the horizon? Will we soon convert the solar system to computronium?
Let’s continue this discussion next Monday–so if you haven’t read the book yet, you still have a whole week to finish.