This lecture was the original inspiration to the mix-algorithm, although not the direct motivation. I consider Alan Watts to be very intelligent, from other material about him, and found his view of "money is just an abstraction" very surprising. I wanted to test this hypothesis by creating an algorithm that doesn't work like traditional economics, but still classify as an economic system: Free trading, controlled inflation, incentivizing rewards. It is not about providing a full working system, but studying a toy model that have some interesting properties. A such toy model should exist if the view of Alan Watts is correct, so I tried to make an algorithm that embodied the ideas Alan suggested, by using negative tax.
Reading through the transcription a year later, I find it full of statements that I earlier disregarded as jokes or cliches, but might have been said with a much more careful thought by Alan Watts. He describes the relation between psychological hang-ups and automation, that the whole purpose of a machine is to free humans from doing repetitive and boring work.
That an economic system can be designed such that it blocks this from happening, is a proof there is something wrong with the system, which I would agree if there was a such blockage. At this point, I think he is a only correct for some cases, since the system still encourages automation, only with the difference that it benefits fewer people initially. However, the idea that meaningless jobs are created to avoid automation is true.
I think his remarks about deciding how to control tokens for exchange while allowing automation is spot on. This seems rather sane, and perhaps we are too busy to notice that our beliefs about money are predating modern technology. That people today basically do nothing to fix it, is a proof of irrational thinking (what he calls "guilt" or "psychological blockage"). Also, he points out the big challenge of changing that way of thinking, which I think he underestimated, since what he proposed would never work today. I believe humans are too indoctrinated to even consider this as a real option, even after repeatedly showing evidence of the opposite. Which makes the idea more interesting to work on, of course.
I am still open to the idea that Alan Watts might be wrong. Personally I am very skeptical to thoughts about modern society is being terribly wrong about economics, given the amount of wealth technology has provided, but I have also seen modern society being terribly wrong in lots of other areas. I can not exclude the possibility that we are also wrong about the fundamental nature of money, and want to check whether this is true for myself.
One thing was that Alan Watts suggested this view in the 60s, possibly from motivations about society, but today it is a much harder problem. We have mathematical proofs that rational agents using logical induction are not exploitable over an upper bound of computational capacity. This raises the bar at an extremely higher level that I believe Alan Watts never anticipated, one that proposes that in the far future our current economic system could be damaging for humanity.
Unlike Alan Watts, my motivation is not about a better society, but primarily preventing a catastrophe, since a better society is the likely outcome if we can survive long enough. I believe the current economic system in the real world encourages artificial agents unnecessarily, exposing humanity for risks involving the control problem, where shared use of narrow artificial intelligence would suffice. There is no efficient wealth distribution mechanism in place, and the system seems vulnerable against increasingly rational agents, with majority of humanity as the loosing part. The solution should be more focus on AI safety and reduction of economic inequality.
So, it seems worth investigating algorithms that distribute wealth when heavily pressured by automation.
http://www.scottsantens.com/yeah-but-who-is-going-to-pay-for-it-basic-income-alan-watts
My thoughts about the lecture:
This lecture was the original inspiration to the mix-algorithm, although not the direct motivation. I consider Alan Watts to be very intelligent, from other material about him, and found his view of "money is just an abstraction" very surprising. I wanted to test this hypothesis by creating an algorithm that doesn't work like traditional economics, but still classify as an economic system: Free trading, controlled inflation, incentivizing rewards. It is not about providing a full working system, but studying a toy model that have some interesting properties. A such toy model should exist if the view of Alan Watts is correct, so I tried to make an algorithm that embodied the ideas Alan suggested, by using negative tax.
Reading through the transcription a year later, I find it full of statements that I earlier disregarded as jokes or cliches, but might have been said with a much more careful thought by Alan Watts. He describes the relation between psychological hang-ups and automation, that the whole purpose of a machine is to free humans from doing repetitive and boring work.
That an economic system can be designed such that it blocks this from happening, is a proof there is something wrong with the system, which I would agree if there was a such blockage. At this point, I think he is a only correct for some cases, since the system still encourages automation, only with the difference that it benefits fewer people initially. However, the idea that meaningless jobs are created to avoid automation is true.
I think his remarks about deciding how to control tokens for exchange while allowing automation is spot on. This seems rather sane, and perhaps we are too busy to notice that our beliefs about money are predating modern technology. That people today basically do nothing to fix it, is a proof of irrational thinking (what he calls "guilt" or "psychological blockage"). Also, he points out the big challenge of changing that way of thinking, which I think he underestimated, since what he proposed would never work today. I believe humans are too indoctrinated to even consider this as a real option, even after repeatedly showing evidence of the opposite. Which makes the idea more interesting to work on, of course.
I am still open to the idea that Alan Watts might be wrong. Personally I am very skeptical to thoughts about modern society is being terribly wrong about economics, given the amount of wealth technology has provided, but I have also seen modern society being terribly wrong in lots of other areas. I can not exclude the possibility that we are also wrong about the fundamental nature of money, and want to check whether this is true for myself.
One thing was that Alan Watts suggested this view in the 60s, possibly from motivations about society, but today it is a much harder problem. We have mathematical proofs that rational agents using logical induction are not exploitable over an upper bound of computational capacity. This raises the bar at an extremely higher level that I believe Alan Watts never anticipated, one that proposes that in the far future our current economic system could be damaging for humanity.
Unlike Alan Watts, my motivation is not about a better society, but primarily preventing a catastrophe, since a better society is the likely outcome if we can survive long enough. I believe the current economic system in the real world encourages artificial agents unnecessarily, exposing humanity for risks involving the control problem, where shared use of narrow artificial intelligence would suffice. There is no efficient wealth distribution mechanism in place, and the system seems vulnerable against increasingly rational agents, with majority of humanity as the loosing part. The solution should be more focus on AI safety and reduction of economic inequality.
So, it seems worth investigating algorithms that distribute wealth when heavily pressured by automation.