"History shows that technological revolutions draw out human potential.”
Well, let’s take a look at the major technological revolutions in human history
1. Tool making. Yes, it gave us enormous advantages right from the start. It also gave us the idea that we could manipulate the world to our design. And how has that turned out? We are approaching the point at which our very existence is in question,.
2. The Agricultural Revolution. Yes, it enabled us to feed far more people at the same time that it freed an increasing majority of us to do other things. How has that turned out? We have flooded the world with ourselves to the extent that a substantial portion of the rest of life on earth is at risk of extinction (except of course that part of the biome that we keep in close captivity in often appalling conditions in order to feed us), and vast numbers of us re packed into cities that are breeding grounds for disease and violence.
3 Writing. Yes, it freed us from the prison of human memory, allowing us to use our accumulating knowledge. How has that turned out? We’ve built both better and better technology which does improve life, and at the same time built weapons capable of ending all life on earth.
4. Steam power – the first invention that significantly enhanced human muscle power in ways that have transformed the world. How has that turned out? We’ve turned much of our atmosphere into an ncreasingly dangerous mixture of pollutants.
5. The computer. Yes, it has given us powers of usage and creation utterly beyond any previous belief. And the internet, well, I won’t comment on the mixed bag that is.
And yes, this has been a simplification that does not do justice to anywnere near all the facts. But this we do know.
Like all the other really significant inventions, AI will turn out to be a double edged sword, because in the final analysis, our tools are as good or as bad, as advantageous or as dangerous as the men and women using them.
This essay suggests why economists need to get out more. The abstraction and generality of the analysis ("we" will do this, "human nature" will do that, etc.) completely ignores the fact that millions of careers will be destroyed and people--real people (not the abstract categories moved about on the chessboard in this piece)--will have their dignity and livelihoods destroyed. For those people, this destruction is forever. There lives will be ruined. "Well, them's the breaks, we'll catch you down the line in the next generation," doesn't somehow add up for me.
In PhD dissertation in 1971, I developed and tested algorithms to determine the content analysis of prose. The correlation of the algorithm's findings and the same task with humans showed a high level of agreement.
My committee got into a fierce argument at my final orals over what it was I was doing. In retrospect, it was AI. Norbert Weiner noted that the lines of science were drawn by people too dumb to understand that there were no lines to science.
About the same time, Lewis R Goldberg, a former colleague of mine, wrote a seminal article entitled "Man vs Model of Man". It showed that a computational model based on on radiologists estimates of the outcome of patients with lung cancer did better than the radiologists. Interestingly, radiologists working with the results of Goldberg's model did better than the model. Alfred North Whitehead noted that advances of the mind are not made by operations of the mind but removing operations from the mind.
Viewed this way, AI is a tool of the mind just like a typesetting machine is a tool of a printer. Likewise, there were predictions when trains first started running that people could not survive the speed they would soon be capable of. Tell that to the pilot of a supersonic airplane. History is replete with examples of experts not understanding new things.
> But this narrative misses a crucial piece of the plot. Even in a world where every task we can name has been taken on by more capable machines, human beings will want and imagine a still better set of circumstances.
In the worried-about scenario, these circumstances will be acheived by having AI achieve them so that humans will still be out of the loop
"History shows that technological revolutions draw out human potential.”
Well, let’s take a look at the major technological revolutions in human history
1. Tool making. Yes, it gave us enormous advantages right from the start. It also gave us the idea that we could manipulate the world to our design. And how has that turned out? We are approaching the point at which our very existence is in question,.
2. The Agricultural Revolution. Yes, it enabled us to feed far more people at the same time that it freed an increasing majority of us to do other things. How has that turned out? We have flooded the world with ourselves to the extent that a substantial portion of the rest of life on earth is at risk of extinction (except of course that part of the biome that we keep in close captivity in often appalling conditions in order to feed us), and vast numbers of us re packed into cities that are breeding grounds for disease and violence.
3 Writing. Yes, it freed us from the prison of human memory, allowing us to use our accumulating knowledge. How has that turned out? We’ve built both better and better technology which does improve life, and at the same time built weapons capable of ending all life on earth.
4. Steam power – the first invention that significantly enhanced human muscle power in ways that have transformed the world. How has that turned out? We’ve turned much of our atmosphere into an ncreasingly dangerous mixture of pollutants.
5. The computer. Yes, it has given us powers of usage and creation utterly beyond any previous belief. And the internet, well, I won’t comment on the mixed bag that is.
And yes, this has been a simplification that does not do justice to anywnere near all the facts. But this we do know.
Like all the other really significant inventions, AI will turn out to be a double edged sword, because in the final analysis, our tools are as good or as bad, as advantageous or as dangerous as the men and women using them.
This essay suggests why economists need to get out more. The abstraction and generality of the analysis ("we" will do this, "human nature" will do that, etc.) completely ignores the fact that millions of careers will be destroyed and people--real people (not the abstract categories moved about on the chessboard in this piece)--will have their dignity and livelihoods destroyed. For those people, this destruction is forever. There lives will be ruined. "Well, them's the breaks, we'll catch you down the line in the next generation," doesn't somehow add up for me.
A rare sane view of AI.
In PhD dissertation in 1971, I developed and tested algorithms to determine the content analysis of prose. The correlation of the algorithm's findings and the same task with humans showed a high level of agreement.
My committee got into a fierce argument at my final orals over what it was I was doing. In retrospect, it was AI. Norbert Weiner noted that the lines of science were drawn by people too dumb to understand that there were no lines to science.
About the same time, Lewis R Goldberg, a former colleague of mine, wrote a seminal article entitled "Man vs Model of Man". It showed that a computational model based on on radiologists estimates of the outcome of patients with lung cancer did better than the radiologists. Interestingly, radiologists working with the results of Goldberg's model did better than the model. Alfred North Whitehead noted that advances of the mind are not made by operations of the mind but removing operations from the mind.
Viewed this way, AI is a tool of the mind just like a typesetting machine is a tool of a printer. Likewise, there were predictions when trains first started running that people could not survive the speed they would soon be capable of. Tell that to the pilot of a supersonic airplane. History is replete with examples of experts not understanding new things.
> But this narrative misses a crucial piece of the plot. Even in a world where every task we can name has been taken on by more capable machines, human beings will want and imagine a still better set of circumstances.
In the worried-about scenario, these circumstances will be acheived by having AI achieve them so that humans will still be out of the loop