8 min read ⌚
The Secret of Human Thought Revealed
Humans are capable of doing so many things computers will never be, right?
Show me a computer capable of thinking, writing symphonies, loving, etc. – and I’ll show you a flying pig.
Don’t put your mouth where your money is, says Ray Kurzweil. Because you will need to breed a whole new race of pigs in a decade or so.
“How to Create a Mind” explains why – and how – computers will start writing symphonies.
Who Should Read “How to Create a Mind”? And Why?
Ray Kurzweil’s predictions comprise the wettest of futurists’ dreams. And even though “How to Create a Mind” doesn’t state anything new of this sort, every futurist and curious SF thinker has already bought this book by now.
The rest should read it to find what all the fuss is about. Because even if you know nothing about AI and neuroscience, this may be a good time to start learning about it.
At least if you believe Ray Kurzweil and this book.
About Ray Kurzweil
Ray Kurzweil is a prize-winning scientist, writer, and futurist.
A winner of MIT’s “Inventor of the Year” prize in 1988, Carnegie Mellon’s top science Dickson Prize six years later and “National Medal of Technology and Innovation” in 1999, Kurzweil has so far received at least 21 honorary doctorates, and special honors from three different U.S. presidents.
He has invented numerous things, ranging from the first omni-font OCR (optical character recognition) to the first print-to-speech reading machine for the blind, from the first flatband scanner to the first commercial text-to-speech synthesizer.
So, you could say that he’s partially responsible for the Siris, Alexas, and Cortanas you talk to on a daily basis.
Unsurprisingly, in 2002, Kurzweil was inducted into the National Inventors Hall of Fame.
He has authored numerous articles and books, including “The Singularity Is Near.”
“How to Create a Mind Summary”
Westworld’s second season premiered last night on HBO.
And we felt that there was no better moment to provide you with a summary of a book titled “How to Create a Mind.”
Especially if it is brought to you by a man who has not only been described by “Forbes” as “the ultimate thinking machine,” but who also has an entire Wikipedia article listing his predictions about the future.
And there’s more where that came from!
Ladies and gentlemen, please join us in unraveling the secret of human thought with the one and only Ray Kurzweil, aka the guy who gave humanity flatbad scanners, optical character recognition, print-to-speech reading machines, and text-to-speech synthesizers!
In a nutshell – someone who definitely knows more than most about how our brain may function, based on his work with artificial brains.
And is there a better way to start a book on thoughts other than with few thought experiments?
Let’s try them out.
First, the simplest: recite the alphabet.
Piece of cake, right?
OK, now recite it backward.
Doesn’t feel as easy, does it?
In fact, chances are, you’re incapable of reciting the alphabet backward no matter how much you try. Even though, if you think about it, you should have no problem: you know all the letters, and you’ve used them thousands and thousands of times.
And, most importantly, you just recited them the other way around!
So, what’s the problem?
We’ll get to that in a second.
But, before, try with us another thought experiment. This time, try to visualize a person you’ve seen only once or twice in your whole life. If you can’t think of any, try thinking about your short trip to the local store this morning.
Can you envisage even one single person of the few you passed by?
No, you can’t.
Kurzweil thinks that these thought experiments reveal something much more than the fact that, essentially, your memory sucks.
Namely, that everybody’s memory sucks in the same way. And that this should give us a hint on how our brain is actually doing its job.
You thought that only computers follow specific algorithms?
Guess again: you do too!
So much so that, in fact, human consciousness pioneer Benjamin Libet has proposed that even your free will may be an illusion!
Since, according to him, these experiments show that your brain is also merely – OK, in strictly relative terms – doing hierarchical statistical analysis.
And by brain, we actually mean your neocortex, which, according to Kurzweil is where the magic actually happens.
We all know that the neocortex is the most advanced part of our brains and is what makes us so different from the rest of the animal world.
Now, according to Kurzweil, this is because the human neocortex contains about 300 million hierarchically arranged general pattern recognizers. And, as the thought experiments we explained above prove, these pattern recognizers aren’t interested in sounds, images, videos, or smells.
The only thing they are interested in is patterns.
That’s why you can’t recite the alphabet backward – it should be easy if your brain remembered information and data. But if your brain remembers patterns, reciting the alphabet back or playing a song from the middle is the same as starting to read a book from page 147.
That’s why you can’t remember people you’ve only seen once or twice in your life as well. In fact, police profilers intuitively know this, so they stimulate the memory of witnesses by showing them different types of eyes, brows, or mouths.
Because, as Marcel Proust taught us, there’s a particular type of memory, involuntary memory, which is triggered once an external stimulus hits the right note of the pattern.
You know what we’re talking about!
You can’t remember a song even though someone is singing the middle part of it. But, then someone sings the right sequence and the middle section falls neatly into place!
Finally, pattern recognition is why all of the memory techniques memory champions advise us to use are pattern-related. And even more – hierarchically ordered.
Now, if your brain works this way – i.e., as if an automat – shouldn’t computer scientists be capable of creating an artificial mind?
Yes, they should.
And in Kurzweil’s opinion – using hidden Markov models and genetic algorithms – they inevitably will by 2029.
Why shouldn’t they?
Intel has already devised a way to trick the limitations of Moore’s law by inventing 3D processors. Japan’s supercomputers are already capable of running 1016 calculations per second – which is just as much as a digital neocortex will need to function.
Finally, the data it should store – around 20 billion bytes (300 million patterns * 72 bytes) amounts to no more than 20 GB, i.e., the size of your USB.
Because, as it has been proven over and over again in the past – whether in science or art – it’s not the amount of data that’s important; it’s the actual and potential interconnections inside it.
So, brace for it – Kurzweil claims that AI humanoids indistinguishable by brain power from humans will become a reality in less than 12 years.
We guess the remaining question at this point is: should you believe Kurzweil?
Well, remember the list with predictions we mentioned at the beginning of this summary? It was made back in 1989. And in October 2010, twenty years later, Kurzweil published a PDF titled “How My Predictions Are Faring.”
In 147 pages, the document lists as many predictions. 12 of them are deemed to be “essentially correct,” 17 “partially correct” and 3 – “wrong.”
As for the rest 115?
Let us write this in all caps because it’s that important:
Ladies and gentlemen, set your watches: we’re about 12 years away from real-life “Westworld.”
For better or for worse, the countdown commences.
Key Lessons from “How to Create a Mind”
1. Pattern Recognition Theory of Mind
2. Welcome to Searle’s Chinese Room: How Do You Know You’re Not a Machine?
3. The Untethered Artificial Mind: The Artificial Mind Which Learns
Pattern Recognition Theory of Mind
How do we think?
Do we think through data, logic, images, sounds, smells?
Neither, says Ray Kurzweil: we think only and exclusively through patterns.
Our neocortex contains about 300 million general pattern recognition circuits which hierarchically structure our memory and experiences.
In other words, if we translate this into practical example (say, how we read), the process looks something like this.
Namely, some of these recognizers are low-level and see only straight and diagonal lines. But, they transmit this information to the higher echelons which are then capable of recognizing letters. These pass on the message to the word-level recognizers, etc. etc.
The information moves back and forth and, based on previous patterns, in time, the recognizers learn to predict the info ahead. That’s how speech recognition works, and that’s why sometimes you see transcribed YouTube captions revealing words before you hear them.
That is your brain as well.
And yes – it gets a bit strange from here on.
Welcome to Searle’s Chinese Room: How Do You Know You’re Not a Machine?
You see, back in 1980, philosopher John Searle made the distinction between weak AI and strong AI based on a simple experiment.
Say you make a program capable of taking Chinese characters as inputs, analyzing them profoundly and giving the expected outcome. And say this program is so convincing that even a Chinese can’t see anything wrong with it and, thus, it passes the Turing test.
The question is: does the program really understands Chinese?
Searle argued against this, by claiming that if he is locked in a room with the machine’s in-programmed manual, and receives the same inputs under the door, he should be able to give the same answers back by merely following the same instructions the machine does.
However, he doesn’t speak a word of Chinese.
Kurzweil says: OK, that may be true.
But what if your brain works the same way?
Let’s not forget that Watson destroyed the best humans in Jeopardy!
The Untethered Artificial Mind: The Artificial Mind Which Learns
It’s time you stopped thinking about machines in terms of programs – unless you start thinking about yourself in the very same way.
In other words, our brains are nothing less – or more – than a pattern recognizing structures. However, this is such a powerful method to acquire new information that it has got us – humans – to a place where we are capable of creating other creatures similar to us.
Because once we perfect a brain capable of recognizing patterns (and we’re already there: think speech recognition), we will essentially create a machine capable of teaching itself. And since a machine’s neocortex can be improved, in time, we will be able to develop machines which will be vastly superior to us.
That’s right: we’re talking about a new species.
Like this summary? We’d like to invite you to download our free 12 min app, for more amazing summaries and audiobooks.
“How to Create a Mind Quotes”In mathematics you don’t understand things. You just get used to them. (via John von Neumann) Click To Tweet We are a pattern that changes slowly but has stability and continuity, even though the stuff constituting the pattern changes quickly. Click To Tweet 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 learning process are encoded in DNA. Click To Tweet Human beings have only a weak ability to process logic, but a very deep core capability of recognizing patterns. To do logical thinking, we need to use the neocortex, which is basically a large pattern recognizer. Click To Tweet Philosophy is a kind of halfway house for questions that have not yet yielded to the scientific method. Click To Tweet
Our Critical Review
“How to Create a Mind” may be uneven and repetitive at times, but, even so, it’s exceptional. Some have deemed its subtitle a bit overpromising, but to others, the book actually manages to give us the most complete theory on how we may think.
Now, if Kurzweil is right about that, then creating an artificial mind is not far ahead. And if that is true, then you reading this book should become a reality in the following weeks.