Not saying we don’t need to fix things… we need to destroy even the concept of billionaires. While things are bad, and trending worse, they’re not yet “literally eat the rich” bad.
Only when not looked at on a global scale (such as 1% owns 99% being assumed to be about the USA rather than global wealth). Mormengil’s opening response is very feels over facts (also the claim that there was no state support for the poor can be technically true, but churches and local elite as well as royal dictat were often involved in poor relief and charity in the Middle Ages), the later response are better detailed.
And in the 1200s, global wealth inequality and access to food was for much of the world better or comparable to where it is now.
But global wealth inequality and access to food got worse after colonisation rearranged American and then African economies for European, and then USAian, benefit.
AI is already filled with implicit bias towards the current status quo. It can be very tricky to get AI chatbots to give anything other than platitudes about inequality, and they’re often very quick to try to shut down or redirect talk of system change.
To think that they’ll not reflect continued post colonial and extractivist systems of power seems, to me, shortsighted.
You would just have to let an superintelligent (aligned) AI robot loose and prompt it to produce enough food for everyone. It wouldn’t even be any maintaining effort, once the robot had been created. If it doesn’t have any negative consequences to the creators to have positive consequences for everyone else, and there are any empathetic people on the board of creators, I don’t see why it wouldn’t be programmed to benefit everyone.
True, and I have my doubts on the alignment problem being solved. But that’s a technical problem, a separate conversation from whether even attempting it is worthwhile in the first place.
Increasing wealth has only ever been observed to fuel greater inequality.
I don’t see any evidence that the value that increasing automation is bringing will be distributed more evenly.
We produce enough food for everyone and still let people starve - equal access to AI is even harder to justify than equal access to food.
I’m not so sure about that. When we compare medieval wealth inequality to now, it was worse back then. Ew, a link to Reddit, but it’s got good info.
Not saying we don’t need to fix things… we need to destroy even the concept of billionaires. While things are bad, and trending worse, they’re not yet “literally eat the rich” bad.
Only when not looked at on a global scale (such as 1% owns 99% being assumed to be about the USA rather than global wealth). Mormengil’s opening response is very feels over facts (also the claim that there was no state support for the poor can be technically true, but churches and local elite as well as royal dictat were often involved in poor relief and charity in the Middle Ages), the later response are better detailed.
And in the 1200s, global wealth inequality and access to food was for much of the world better or comparable to where it is now.
But global wealth inequality and access to food got worse after colonisation rearranged American and then African economies for European, and then USAian, benefit.
AI is already filled with implicit bias towards the current status quo. It can be very tricky to get AI chatbots to give anything other than platitudes about inequality, and they’re often very quick to try to shut down or redirect talk of system change. To think that they’ll not reflect continued post colonial and extractivist systems of power seems, to me, shortsighted.
You would just have to let an superintelligent (aligned) AI robot loose and prompt it to produce enough food for everyone. It wouldn’t even be any maintaining effort, once the robot had been created. If it doesn’t have any negative consequences to the creators to have positive consequences for everyone else, and there are any empathetic people on the board of creators, I don’t see why it wouldn’t be programmed to benefit everyone.
As long as it doesn’t generate any negative externalities, sure. That’s a huge alignment problem though.
True, and I have my doubts on the alignment problem being solved. But that’s a technical problem, a separate conversation from whether even attempting it is worthwhile in the first place.