It began with a fairly innocuous purchase - that of Netflix, for a reason so simple now that it is not worth mentioning. And it was as we sat around fat and stuffed and bored on Thanksgiving night that we explored its possibilities. And it was I – I believe it was I – who noticed that they offered Sherlock and I said, “The Sherlock Holmes series! Jim told me about that. Let’s give it a try.” It is now Tuesday morning and we must have watched 15 hours of Sherlock since that fateful day. I would have watched more if my wife had not turned off the system at 11:30 PM saying, “that’s enough!” I was mad then, but grateful now.
I had loved the Sherlock Holmes stories as a middle-schooler, trying to be as smart as him then, and I have found that now I am doing the same thing. I want to be brilliant like Sherlock. I want to let public opinion and normal superficial social niceties, vagaries, and outright lies fall by the wayside as things too small and petty for such a great mind as mine. And of course, I want to win. Yes, I still want to be Sherlock. I am still a nerd at heart and it is apparent that none of life’s tough lessons have really pried me from nerd pride. Ten pounds heavier after Thanksgiving and wheezing from the weight of it, I have met myself again. But the pride is now dropping away, for I am beginning to understand that Sherlock isn’t so smart after all, even if he is way, way smarter than I.
For one thing, he is a fictional character who is supposed to be a genius, but is created by a writer or writers, who are not geniuses, at least in the Sherlock way. In other words, they make up a lot of stuff to make him look like a genius. But there is much more and less to it than that.
Elon Musk stated five years ago that we would be taken over by AI (artificial intelligence) robots within five to ten years. That has not happened, but he is insistent that it will happen soon, certainly within a generation. He has said that it is already too late to stop it – that AI is so much smarter than we are that it will pass us by, and eventually eliminate us, as HAL almost did in “2001 – A Space Odyssey.” The only solution, he claims, is to become AI ourselves – that is, to insert an artificial neural web into our brains so that we can get immediate and unlimited information, too, just like our robots.
Now, Musk might be a genius in his own way, but this is the stupidest idea I have heard in a long time, for two reasons: one - the more obvious one - that by inserting computers into our brain we will open ourselves up to be controlled by computers in such a way that the minions of North Korea would look like free- thinking Bohemian artists in comparison. Horrible idea, for that alone. More important, however, is reason number 2, dealing with the nature of the information itself.
In the series with Sherlock, when he is putting together his “deductions” we are shown graphics that appear in his mind as if he were a computer. His thinking is lightning fast and leaves no detail unturned. He can figure out nearly every situation or person in no time at all – including what one’s habits are and who is sleeping with whom. But just as Sherlock is really made by a normal IQ writer, so these “genius” AI robots are made by people only using one element of intelligence. Simply put, the programmers can’t transfer the depth of their own programming into the robots, but only the perception of that intelligence from the outside – like someone explaining language from a grammar book alone, rather than presenting it truly in all its social subtleties as well as from the un-relatable unconscious level. Thus, the robots are and will always be LESS than their creators, no matter how fast they can compute.
And more. The AI people would argue that their robots are programmed to learn, and thus could overcome the problem mentioned above, just as a human would. I doubt it, but let’s go with that. Still, the robots could never think outside the capability of their software (and their hardware, which is also based on specific concepts). This software was made for strictly rational purposes. It does not carry the capacity to appreciate beauty, or wonder, or awe or gratitude. It might be taught to simulate it, but it could never have it, inside; that is, it could never have a soul. Leaving besides the deeper spiritual repercussions, it could therefore never get inspiration, which comes from ----well, nowhere, as far as our rational mind is concerned. Rather, inspiration, true genius, comes from the same area (for lack of a better word) as does wonder, awe and beauty. And even these are subsumed by all the possibilities that make up the human potential, about whose source we haven’t a clue. Since we cannot even imagine that source, we certainly cannot manage that source for programming.
Robots, then, would have an extremely limited reality base, one that must abide by our own very limited concept of reason – that is, by the type of thought that we can use to manipulate our current concept of the universe - which certainly is not the ultimate concept of the universe.
In more poetic and spiritual words – and more accurate ones – the robots would not only have no soul, but they would have no genius, no voice from nowhere that could leap across analytical thinking to bring absolutely novel ideas based on absolutely novel reasoning. While the robots would, like Sherlock, be so much faster and more efficient than we are in our normal world, they could not transcend it. They would be much better at it, but still, they could not transcend it. We, on the other hand, could and probably will have new insights that will make such AI obsolete – and in a fairly short period of time.
That does not mean that AI is without risks. It can be programmed to operate as soldiers, and be programmed with the artificial desire – that is, the coded aim – to dominate the world. But it would operate on a premise about reality that is not whole and that is static within a certain plane of possibilities. If it did not dominate us fast, we would kill it; we would eventually outsmart it and win.
But here I go again, trying to be Sherlock with all my fancy reasoning, which is so difficult to present verbally exactly because our understanding can take us beyond normal reasoning. Rather, perhaps it is simply enough to know that we are programmed by a designer who can not only make the universe, but all the laws thereof, and all the potential thoughts and concepts ever. We could never be that designer; even at our very best, we could never top the works of the original designer – which include us. Musk, I’m afraid, has fallen victim to the sin of pride – or at least has been trapped by his own conditioning to such an extent that he cannot see how simple it really is. Elementary, my dear Watson. FK