Header 1

Our future, our universe, and other weighty topics


Thursday, March 22, 2018

The Not-So-Grand Destiny of Robots

It finally happened. A robot car killed a pedestrian. A woman crossing the street was killed by a self-driving car. It seems rather amazing that such a thing didn't happen earlier with a self driving car. That's because robots don't understand a single thing.

You may have heard of “machine learning.” But there is absolutely no such thing as machine understanding. There is not a robot or computer in the world that understands anything. Understanding is something that goes on in a mind. Robots and computers don't have minds.

You may get a clearer idea of how things are for a self-driving car if you consider that such a car has not the slightest grasp of such concepts as what a human being is, what life is, or what death is. If you walk in front of some self-driving car, to that car you are merely a group of pixels to be avoided. To the self-driving car, you are only slightly different from stop signs or electrical poles. The software of the self-driving car probably has some priority system whereby it assigns a higher value to avoiding collisions with mobile objects than stationary objects. But the car has not the slightest understanding of why it is bad to run over pedestrians. 

 

Let us look into what happens when computers and robots compute. The following equation covers most of the types of computation that occur.

digital inputs + processing = digital outputs or modification of digital data

There are various types of variations of this equation. One is simply:

no inputs + processing = digital outputs or modification of digital data

Another variation is the following:

digital inputs + processing + retrieval of other digital inputs = digital outputs or modification of digital data

By digital inputs or digital outputs I mean anything at all that can be represented digitally, by a sequence of binary numbers. Here are some of the things that we know can be represented digitally, and which modern computers do use as digital inputs or digital outputs:

Any number
Any set of characters or words
Images
Sound recordings
Videos
Databases

Any text can be digitally represented by means of things such as the ASCII system that allows you to represent particular characters as particular numbers. While we don't normally think of an image as digital, it can be represented digitally as a series of pixels or picture elements. For example, a photograph might consist of 1 million pixels, which each can be represented by a number representing a particular shade of color. So the image can be digitally represented by a million such numbers. A video or real-time camera input can also be representing digitally, since the video can be represented as a series of images, each of which can be digitally represented.

Is the output of a self-driving car a digital output? It is. The car simply computes 5 digital quantities:

  • how much to turn the steering wheel to the left;
  • how much to turn the steering wheel to the right;
  • how much to accelerate;
  • how much to brake;
  • which of three positions the transmission should be in (drive, park or reverse)

After a digital-to-analog conversion, these digital quantities are applied to the car's controls.

But there are some things that we can never hope to produce as digital outputs. The first is real conceptual understanding. By understanding I don't mean “how-to” type understanding, but the high-level conscious understanding of some abstract truth or concept. We can imagine no possible way to produce a digital output that would equal a real conceptual understanding of something.

But, you may ask, doesn't that smart computer Watson already understand something – the game of chess? No, it doesn't. Watson merely can produce a digital output corresponding to a good move to make as the next move in a chess game. Watson has zero conceptual understanding of the game of chess itself, and has zero understanding of the abstract concept of a game. The only way you can understand the abstract concept of a game (or the abstract concept of leisure) is if you have been a human being (or something like a human), and played a game yourself.

A digital output must always boil down to a series of 1's and 0's. Can we imagine a series of 1's and 0's that would equal a real understanding of an abstract concept such as health, matter, life or world peace? No, we cannot. Understanding is not a digital output.

So none of our computers and none of our robots understand anything. A programmer may say, “My program understands the difference between new customers and old customers,” but that's just a kind of loose way to talk. What the programmer really means is that the logic in his program can distinguish between two different things. That's not the same thing as actual understanding. Here “distinguish” simply means “behave differently based on different situations or inputs.”

Basically all of the computers and robots existing today are digital. No one has ever built a computer or robot that operates like the human brain operates.

There is also an extremely strong case for believing that it will never work to try to give a computer or robot understanding by
creating a computer or robot based on the human brain. The case is based on reasons for doubting that the human brain is actually the source of the human mind.

The claim that the human mind is produced by the human brain is something scientists have never established. Such a claim is merely a speech custom of scientists, an unproven dogma. There are some very good reasons for doubting this claim, which scientists offer as a simplistic explanation because they lack the imagination to ponder more profound possibilities. Below are a few (but by no means all) of the many reasons for doubting the “brains produce minds” claim:
  • the fact that there are many dramatic cases in the medical literature of people who had more or less normal minds even though large fractions of the brain (or most of their brains) were destroyed due to injury or disease, including super-dramatic cases of people with good minds but less than 15 percent of their brains;
  • the fact that there is no scientific understanding at all of how brains or neurons could be producing consciousness, thought, understanding or abstract ideas (mental things that are very hard or impossible to explain as coming from physical things);
  • the fact that there is no plausible account to be told of how brains could possibly be storing memories that last for fifty years, given the high protein turnover in synapses, where the average protein only lasts a few weeks;
  • the fact that there is no understanding of how brains could achieve the instantaneous recall of distant, obscure memories that humans routinely show, given the lack of any coordinate system or indexing in a brain that might allow some exact position of a stored memory to be very quickly found;
  • the fact that there is no understanding whatsoever of how concepts, visual information, long series of words, and episodic memories could ever be physically stored by a brain in any way that would translate all these diverse types of information into synapse states or neuron states;
  • the fact that for more than 40 years numerous people have reported vivid near-death experiences occurring after their hearts stopped and their brains were inactive, during times when they had no brain waves, and they should have had no consciousness at all, with many of the medical details they reported during such experiences being independently verified (as described here).

All of these reasons (and quite a few others I have not listed) strongly suggest that the human mind comes from something other than the human brain. Rather than being a bottom-up phenomenon bubbling up from tiny little neuron events, our minds may come to us through a top-down process, arising from some mysterious cosmic reality that we don't understand – perhaps something along the lines suggested here and here.

What is the robotic relevance of the evidence strongly suggesting that our minds do not come from brains? Such evidence strongly suggests that all future attempts to build brilliant robots by imitating the architecture of the brain will fail. If we're not getting our minds from brains, there is no reason at all to think that robots or computers will ever get real conceptual understanding because robot designers or computer designers will one day use some architecture based on the human brain.

If that is so, we need not fear at all that robots or computers will ever take over the world, because they will always lack the understanding to do such a thing.  The destiny of robots will not be the grand destiny of taking over as masters of the planet, but the much more modest destiny of serving as our slaves, servants and toys.

Postscript: For a collection of posts presenting in detail many aspects of the case for thinking that minds do not come from brains, see this site.