1:38 AM
OGB community
Recently, I've become aware that various authors and researchers are predicting an event in human history that will change everything we label "human."
Crossing this threshold will allow us to do two things-build machines that are billions of times smarter than we are, and radically increase the lifespan of the individual.
The first objective will be achieved by plunging ahead in the development of computers and artificial intelligence, so that these machines will, in turn, invent greater machines in a quickening arc.
The second objective will arrive as we utilize genetic manipulation, nanotechnology, and "human parts" replacement.
Let me focus on the first objective in this article. And I'll start here: Smartness, intelligence, brilliance, mental capacity, etc. are all based on what?
They are based on the notion that solving problems can be vastly speeded up and made more
effective-and the problems being referred to are those which "the whole human community" shares. War, hunger, pollution, tribal and national conflicts, diminishing supplies of natural resources, and so on.
Here is the central point, however. Regardless of the level of IQ and the speed of reasoning, a problem is a problem is a problem. In other words, any solution depends upon assumptions about what your (our) goals are. "Greatest good for the greatest number," for example, means nothing unless the machine solving a problem operates according to specified priorities that depict and define "greatest good." Without that, a machine is lost. It just sits there and does nothing.
We have to realize there is nothing inherently magical about a machine when it comes to solving problems. A machine isn't suddenly going to "breathe life into itself" so it becomes more capable of setting the most basic goals.
You might recall an old science-fiction movie, "Colossus: The Forbin Project." Two super-computers, one for the USSR and one for the US, are built to assure victory in a nuclear war. Each machine protects itself (by design), so it can't be unplugged. Suddenly, on the brink of war, the machines begin talking to each other and decide the human race is stupid and irretrievably self-destructive. The machines make a pact to protect planet Earth-and essentially recreate it as a world devoted to right-thinking machines, with humans operating as slaves
What's left out of the movie is this: Those computers would never have taken their radical actions "on behalf of the planet," unless humans had inserted relevant goals into their programmed guts.
We are not dealing with some mystical capacity that machines can suddenly attain because of their calculating power.
We are, in fact, dealing with a more sophisticated version of Central Planning. We have seen many societies try this, and we have seen them fail. To turn over all allocation of natural resources and survival decisions to machines could bring on a radically different Era for humans-but not because the machines are better INVENTORS of proper goals for the uman race.
From the point of view of a machine, there are no better or worse goals. There are only those goals which have been programmed into the machines by humans.
As a crass illustration, suppose a machine is given the mandate to solve the climate crisis for the planet. The crisis is defined by scientists through the assumption that global warming is a real and advancing problem that threatens our very existence. Well, machines will then take many actions to solve warming-whether or not it actually exists. And if global warming does not exist at a significant threat level, the machines will perform the most stupid actions imaginable.
Some people object to this "simplistic" analysis. They say, "You have no idea what innovations machines with IQs of 5000 will produce." Actually, I believe I do. They will generate ideas and rules and other machines in line with whatever overall goals and first assumptions are programmed into them. And wherever such assumptions are missing, the machines will fall silent and sit on idle.
Let's try what some might call a best-case scenario. A gaggle of exceedingly capable computers devises a genetically engineered food crop that has astonishing nutritional value and no negative-health downside. The food crop imparts all nutritional needs to humans. It can be grown in a surprisingly small area, because just a few bites from the leaves or fruit are sufficient daily intake for every bodily need.
Next question: Do the machines calculate and put into effect, with the help of other machines, this agri-program for the whole human race or just a limited number of people?
The answer to that depends on the basic assumptions about survival of the species that have been inserted into the machines' thinking apparatus. It could go either way. Some method for such a choice must already exist in the machine-not because the machine is "so smart" it can come to a conclusion on its own, but because it has been given prior direction.
Let us imagine the machine decides to feed half the world's population and force the other half to die, because the planet should only support three billion people. Where did that judgment come from? On what basis was it rendered?
I believe the answer is obvious. The machine contains certain prejudices that have been put there by human programmers.
There is nothing amazing about it. What is amazing is the willingness of technical people to assume that some version of machine IQ, rising to artificial heights, will thereafter produce VALUE-based choices intrinsically more brilliant than anything we poor humans can come up with.
The operative word here is "brilliant," and the fallacy comes about by asserting that the word has something to do with the choice of fundamental values that determine how we run our affairs. That's patently false.
The "rise of the machines" as an ultimate solution for the human race is much like the proposition that a ruling priesthood is much smarter than the "lower" population. For Europe, you could translate "priesthood" into "divine right of kings."
Put in gross terms, this great New Age allows a ruling elite and its machine surrogates to announce to the human race: "We know what you need and we're going to give it you, so shut up and keep walking down the road and obey the signs and focus your eyes straight ahead.
Crossing this threshold will allow us to do two things-build machines that are billions of times smarter than we are, and radically increase the lifespan of the individual.
The first objective will be achieved by plunging ahead in the development of computers and artificial intelligence, so that these machines will, in turn, invent greater machines in a quickening arc.
The second objective will arrive as we utilize genetic manipulation, nanotechnology, and "human parts" replacement.
Let me focus on the first objective in this article. And I'll start here: Smartness, intelligence, brilliance, mental capacity, etc. are all based on what?
They are based on the notion that solving problems can be vastly speeded up and made more
effective-and the problems being referred to are those which "the whole human community" shares. War, hunger, pollution, tribal and national conflicts, diminishing supplies of natural resources, and so on.
Here is the central point, however. Regardless of the level of IQ and the speed of reasoning, a problem is a problem is a problem. In other words, any solution depends upon assumptions about what your (our) goals are. "Greatest good for the greatest number," for example, means nothing unless the machine solving a problem operates according to specified priorities that depict and define "greatest good." Without that, a machine is lost. It just sits there and does nothing.
We have to realize there is nothing inherently magical about a machine when it comes to solving problems. A machine isn't suddenly going to "breathe life into itself" so it becomes more capable of setting the most basic goals.
You might recall an old science-fiction movie, "Colossus: The Forbin Project." Two super-computers, one for the USSR and one for the US, are built to assure victory in a nuclear war. Each machine protects itself (by design), so it can't be unplugged. Suddenly, on the brink of war, the machines begin talking to each other and decide the human race is stupid and irretrievably self-destructive. The machines make a pact to protect planet Earth-and essentially recreate it as a world devoted to right-thinking machines, with humans operating as slaves
What's left out of the movie is this: Those computers would never have taken their radical actions "on behalf of the planet," unless humans had inserted relevant goals into their programmed guts.
We are not dealing with some mystical capacity that machines can suddenly attain because of their calculating power.
We are, in fact, dealing with a more sophisticated version of Central Planning. We have seen many societies try this, and we have seen them fail. To turn over all allocation of natural resources and survival decisions to machines could bring on a radically different Era for humans-but not because the machines are better INVENTORS of proper goals for the uman race.
From the point of view of a machine, there are no better or worse goals. There are only those goals which have been programmed into the machines by humans.
As a crass illustration, suppose a machine is given the mandate to solve the climate crisis for the planet. The crisis is defined by scientists through the assumption that global warming is a real and advancing problem that threatens our very existence. Well, machines will then take many actions to solve warming-whether or not it actually exists. And if global warming does not exist at a significant threat level, the machines will perform the most stupid actions imaginable.
Some people object to this "simplistic" analysis. They say, "You have no idea what innovations machines with IQs of 5000 will produce." Actually, I believe I do. They will generate ideas and rules and other machines in line with whatever overall goals and first assumptions are programmed into them. And wherever such assumptions are missing, the machines will fall silent and sit on idle.
Let's try what some might call a best-case scenario. A gaggle of exceedingly capable computers devises a genetically engineered food crop that has astonishing nutritional value and no negative-health downside. The food crop imparts all nutritional needs to humans. It can be grown in a surprisingly small area, because just a few bites from the leaves or fruit are sufficient daily intake for every bodily need.
Next question: Do the machines calculate and put into effect, with the help of other machines, this agri-program for the whole human race or just a limited number of people?
The answer to that depends on the basic assumptions about survival of the species that have been inserted into the machines' thinking apparatus. It could go either way. Some method for such a choice must already exist in the machine-not because the machine is "so smart" it can come to a conclusion on its own, but because it has been given prior direction.
Let us imagine the machine decides to feed half the world's population and force the other half to die, because the planet should only support three billion people. Where did that judgment come from? On what basis was it rendered?
I believe the answer is obvious. The machine contains certain prejudices that have been put there by human programmers.
There is nothing amazing about it. What is amazing is the willingness of technical people to assume that some version of machine IQ, rising to artificial heights, will thereafter produce VALUE-based choices intrinsically more brilliant than anything we poor humans can come up with.
The operative word here is "brilliant," and the fallacy comes about by asserting that the word has something to do with the choice of fundamental values that determine how we run our affairs. That's patently false.
The "rise of the machines" as an ultimate solution for the human race is much like the proposition that a ruling priesthood is much smarter than the "lower" population. For Europe, you could translate "priesthood" into "divine right of kings."
Put in gross terms, this great New Age allows a ruling elite and its machine surrogates to announce to the human race: "We know what you need and we're going to give it you, so shut up and keep walking down the road and obey the signs and focus your eyes straight ahead.
Posts
0 comments:
Post a Comment