I would guess that the range of things people propose for the shell game is tractable to get a good survey of. It’d be interesting to try to plot out the system as a causal graph with recurrence so one can point to, “hey look, this kind of component is present in a lot of places”, and see if one can get that causal graph visualization to show enough that it starts to feel clear to people why this is a problem. I doubt I’ll get to this, but if I play with this, I might try to visualize it [edit: probably with the help of a skilled human visual artist to make the whole chart into an evocative comic] with arrays of arrows vaguely like,
a -> b -> c_1 -> c_1
... -> ...
c_n -> c_n
|
v
d_1 ... d_n
^
| | /
v v
f <- e
where c might be, idk, people’s bank accounts or something, d might be people’s job decisions, e might be an action by some single person, etc. there’s a lot of complexity in the world, but it’s finite, and not obviously beyond us to display the major interactions. being able to point to the graph and say “I think there are arrows missing here” seems like it might be helpful. it should feel like, when one looks at the part of the causal graph that contains ones’ own behavior, “oh yeah, that’s pretty much got all the things I interact with in at least an abstract form that seems to capture most of what goes on for me”, and that should be generally true for basically anyone with meaningful influence on the world.
ideally then this could be a simulation that can be visualized as a steppable system. I’ve seen people make sim visualizations for public consumption—https://ncase.me/, https://www.youtube.com/@PrimerBlobs—it doesn’t exactly look trivial to do, but it seems like it’d allow people to grok the edges of normality better to see normality generated by a thing that has grounding, and then see that thing in another, intuitively-possible parameter setup. It’d help a lot with people who are used to thinking about only one part of a system.
But of course trying to simulate abstracted versions of a large fraction of what goes on on earth sounds like it’s only maybe at the edge of tractability for a team of humans with AI assistance, at best.
I would guess that the range of things people propose for the shell game is tractable to get a good survey of. It’d be interesting to try to plot out the system as a causal graph with recurrence so one can point to, “hey look, this kind of component is present in a lot of places”, and see if one can get that causal graph visualization to show enough that it starts to feel clear to people why this is a problem. I doubt I’ll get to this, but if I play with this, I might try to visualize it [edit: probably with the help of a skilled human visual artist to make the whole chart into an evocative comic] with arrays of arrows vaguely like,
where c might be, idk, people’s bank accounts or something, d might be people’s job decisions, e might be an action by some single person, etc. there’s a lot of complexity in the world, but it’s finite, and not obviously beyond us to display the major interactions. being able to point to the graph and say “I think there are arrows missing here” seems like it might be helpful. it should feel like, when one looks at the part of the causal graph that contains ones’ own behavior, “oh yeah, that’s pretty much got all the things I interact with in at least an abstract form that seems to capture most of what goes on for me”, and that should be generally true for basically anyone with meaningful influence on the world.
ideally then this could be a simulation that can be visualized as a steppable system. I’ve seen people make sim visualizations for public consumption—https://ncase.me/, https://www.youtube.com/@PrimerBlobs—it doesn’t exactly look trivial to do, but it seems like it’d allow people to grok the edges of normality better to see normality generated by a thing that has grounding, and then see that thing in another, intuitively-possible parameter setup. It’d help a lot with people who are used to thinking about only one part of a system.
But of course trying to simulate abstracted versions of a large fraction of what goes on on earth sounds like it’s only maybe at the edge of tractability for a team of humans with AI assistance, at best.