From politics: there is an absolutely pivotal issue which EA and STEM types tend to be oblivious to. This is the role of values and ideology in defining what is possible.
For example, it’s obvious in a vacuum that the miracles of automation could allow all humans to live free of poverty, and even free of the need to work. …until conservative ideology enters the picture, that is.
When my conservative mother talks about AI, she doesn’t express excitement that machine-generated wealth could rapidly end poverty, disease, and such. She expresses fear that AI could leave everyone to starve without jobs.
Why? Because granting everyone rights to the machine-generated wealth would be anti-capitalist. Because solving the problem would, by definition, be anti-capitalist. It would deny capitalists the returns on their investments which conservatives regard as the source of all prosperity.
To a conservative, redistribution is anathema because prosperity comes from those who own wealth, and the wealthiest people have proven that they should be trusted to control that wealth because they’ve demonstrated the competence necessary to hoard so much wealth so effectively.
Meanwhile, redistribution would transfer power from the worthy to the unworthy, and thereby violate the hierarchy from which conservatives believe all prosperity springs.
It’s circular logic: the rich deserve wealth and other forms of power because they control wealth, and the poor don’t deserve wealth because they don’t own wealth. This isn’t a conclusion reached through logic—it’s a conclusion reached through a combination of 1) rationalized greed and 2) bombardment with conservative media.
To a hardened conservative, this belief that existing hierarchies are inherently just and valuable is the core belief which all other beliefs are formed in service of. All those other beliefs retcon reality into perceived alignment with this central delusion.
This is also why conservatives think so highly of charity (as opposed to mandatory redistribution): charity redistributes wealth only at the voluntary discretion of the wealthy, and grants the power to allocate wealth in proportion to how much wealth a person owns. To a conservative, this is obviously the best possible outcome, because wealth will be allocated according to the sharp business sense of the individuals who have proven most worthy of the responsibility.
Of course, in practice, power serves itself, and the powerful routinely exploit their wealth to manufacture mass cultural delusions in service of their greed. See climate denial, crypto hype, trickle-down economics, the marketing of fossil gas as a “clean” “transition fuel”, the tobacco industry’s war on truth, the promotion of electric cars over public transport that could actually reduce energy consumption, the myth of conservative “fiscal responsibility” after the debt was blown up by both Reagan and Trump (and Mulroney here in Canada), the framing of conservative policy as “pro-growth” as if postwar high-tax policy didn’t bring about rapid economic expansion and public prosperity, the Great Barrington Declaration and other anti-science pandemic propaganda efforts, and the endless stream of money poured into “free-market” “think tank” corporate propaganda outlets. Of course, there are countless other examples, but I’ll stop there.
If we solve alignment but leave conservatives with the power to command AI to do whatever they want, then AI won’t be used for the benefit of all. Instead, it will be exploited by those who own the legal rights to the tech’s output. And all our alignment work will be for nothing, or next to nothing.
Obviously, then, we must redesign our political systems to value human (or sapient beings’) rights over property rights—a project as inherently progressive and anti-conservative as EA itself. The alternative is corporate totalitarianism.
From politics: there is an absolutely pivotal issue which EA and STEM types tend to be oblivious to. This is the role of values and ideology in defining what is possible.
For example, it’s obvious in a vacuum that the miracles of automation could allow all humans to live free of poverty, and even free of the need to work. …until conservative ideology enters the picture, that is.
When my conservative mother talks about AI, she doesn’t express excitement that machine-generated wealth could rapidly end poverty, disease, and such. She expresses fear that AI could leave everyone to starve without jobs.
Why? Because granting everyone rights to the machine-generated wealth would be anti-capitalist. Because solving the problem would, by definition, be anti-capitalist. It would deny capitalists the returns on their investments which conservatives regard as the source of all prosperity.
To a conservative, redistribution is anathema because prosperity comes from those who own wealth, and the wealthiest people have proven that they should be trusted to control that wealth because they’ve demonstrated the competence necessary to hoard so much wealth so effectively.
Meanwhile, redistribution would transfer power from the worthy to the unworthy, and thereby violate the hierarchy from which conservatives believe all prosperity springs.
It’s circular logic: the rich deserve wealth and other forms of power because they control wealth, and the poor don’t deserve wealth because they don’t own wealth. This isn’t a conclusion reached through logic—it’s a conclusion reached through a combination of 1) rationalized greed and 2) bombardment with conservative media.
To a hardened conservative, this belief that existing hierarchies are inherently just and valuable is the core belief which all other beliefs are formed in service of. All those other beliefs retcon reality into perceived alignment with this central delusion.
This is also why conservatives think so highly of charity (as opposed to mandatory redistribution): charity redistributes wealth only at the voluntary discretion of the wealthy, and grants the power to allocate wealth in proportion to how much wealth a person owns. To a conservative, this is obviously the best possible outcome, because wealth will be allocated according to the sharp business sense of the individuals who have proven most worthy of the responsibility.
Of course, in practice, power serves itself, and the powerful routinely exploit their wealth to manufacture mass cultural delusions in service of their greed. See climate denial, crypto hype, trickle-down economics, the marketing of fossil gas as a “clean” “transition fuel”, the tobacco industry’s war on truth, the promotion of electric cars over public transport that could actually reduce energy consumption, the myth of conservative “fiscal responsibility” after the debt was blown up by both Reagan and Trump (and Mulroney here in Canada), the framing of conservative policy as “pro-growth” as if postwar high-tax policy didn’t bring about rapid economic expansion and public prosperity, the Great Barrington Declaration and other anti-science pandemic propaganda efforts, and the endless stream of money poured into “free-market” “think tank” corporate propaganda outlets. Of course, there are countless other examples, but I’ll stop there.
If we solve alignment but leave conservatives with the power to command AI to do whatever they want, then AI won’t be used for the benefit of all. Instead, it will be exploited by those who own the legal rights to the tech’s output. And all our alignment work will be for nothing, or next to nothing.
Obviously, then, we must redesign our political systems to value human (or sapient beings’) rights over property rights—a project as inherently progressive and anti-conservative as EA itself. The alternative is corporate totalitarianism.