Thank you for the article! I’ve been skeptical of the general arguments for progress (from a LT perspective) for several years but never managed to articulate a model quite as simple/clear as yours.
For instance, they might be able to lengthen our future by bringing forward the moment we develop technologies to protect us from natural threats to our survival. This is an intriguing idea, but there are challenges to getting it to work: especially since there isn’t actually much natural extinction risk for progress to reduce, whereas progress appears to be introducing larger anthropogenic risks ... There might be other claims about the value of progress that are unaffected. For example, these considerations don’t directly undermine the argument that progress is better than stasis, and thus that if progress is fragile, we need to protect it. That may be true even if humanity does eventually bring about its own end, and even if our progress brings it about sooner.
When I’ve had these debates with people before, the most compelling argument for me (other than progress being good under a wide range of commonsensical assumptions + moral/epistemic uncertainty) is a combination of these arguments plus a few subtleties. It goes something like this:
Progress isn’t guaranteed
It’s not actually very plausible/viable to maintain a society that’s ~flat, in practice you’re either going forwards or backwards.
Alternatively, people paint a Red Queen story where work towards advancement is necessary to fight the natural forces of decline. If you (speaking broadly) don’t meaningfully advance, civilizational decline will set in, likely in the span of mere decades or centuries
Sometimes people point to specific issues, eg institiutional rot in specific societies. Or global demographic decline which need to be countered by either greater per-capita productivity or some other way of creating more minds (most saliently via AGI, though maybe you can also do artificial wombs or something)
Regression makes humanity more vulnerable to both exogenous risks (supervolcanoes etc) and endogenous risks (superweapon wars, mass ideological capture by suicidal memes, fertility crisis)
(relatedly, subarguments here about why rebuilding is not guaranteed)
Therefore we need to advance society (technologically, economically, maybe in other ways) to survive, at least until the point where we have a stable society that doesn’t need to keep advancing to stay safe.
This set of arguments only establishes that undifferentiated progress is better than no progress. They do not by themselves directly argue against differential technological progress. However, people who ~roughly believe the above set of arguments will probably say that differential technological progress (or differential progress in general) is a good idea in theory but not realistic except for a handful of exceptions (like banning certain forms of gain-of-function research in virology). For example, they might point to Hayekian difficulties with central planning and argue that in many ways differential technological progress is even more difficult than the traditional issues that attempted central planners have faced.
On balance, I did not find their arguments overall convincing but it’s pretty subtle and I don’t currently think the case against “undifferentiated progress is exceptionally good” is a slam dunk[1], like I did a few years ago.
In absolute terms. I think there’s a stronger case that marginal work on preventing x-risk is much more valuable in expectation. Though that case is not extremely robust because of similar sign-error issues that plague attempted x-risk prevention work.
I think more important than the Hayekian difficulties of planning is that, to my understanding of how the term progress is normally used, part of the reason why all these notions of economic-technological, values, and institutional progress have been linked together under the term “progress” is because they are mutually self-reinforcing. Economic growth and peace and egalitarian, non-hierarchical values/psychology → democracy → economic growth and peace and egalitarian values, a-la Kant’s perpetual peace. You can’t easily make one go faster than the others because they all come together in a virtuous circle.
Further, these notions of progress are associated with a very particular vision of society put forward by liberals (the notion of long-run general progress, not just a specific type of moral progress from this particular law, is an idea mostly reserved to liberals). Progress, then, isn’t moving forward on humanity’s timeline, but moving forward on the desired liberal timeline of history, with a set of self-reinforcing factors (economic growth, democratization, liberal values) that lock-in the triumph of liberalism to eventually usher in the end of history.
I think this understanding makes the notion of undifferentiated progress much more compelling. It is hard to separate the different factors of progress, so differentiating is unreasonable, but progress is also a specific political vision that can be countered, not some pre-determined inevitable timeline. We live at a time where it feels uncontestable (Fukuyama a couple decades ago), but it isn’t quite.
Thank you for the article! I’ve been skeptical of the general arguments for progress (from a LT perspective) for several years but never managed to articulate a model quite as simple/clear as yours.
When I’ve had these debates with people before, the most compelling argument for me (other than progress being good under a wide range of commonsensical assumptions + moral/epistemic uncertainty) is a combination of these arguments plus a few subtleties. It goes something like this:
Progress isn’t guaranteed
It’s not actually very plausible/viable to maintain a society that’s ~flat, in practice you’re either going forwards or backwards.
Alternatively, people paint a Red Queen story where work towards advancement is necessary to fight the natural forces of decline. If you (speaking broadly) don’t meaningfully advance, civilizational decline will set in, likely in the span of mere decades or centuries
Sometimes people point to specific issues, eg institiutional rot in specific societies. Or global demographic decline which need to be countered by either greater per-capita productivity or some other way of creating more minds (most saliently via AGI, though maybe you can also do artificial wombs or something)
Regression makes humanity more vulnerable to both exogenous risks (supervolcanoes etc) and endogenous risks (superweapon wars, mass ideological capture by suicidal memes, fertility crisis)
(relatedly, subarguments here about why rebuilding is not guaranteed)
Therefore we need to advance society (technologically, economically, maybe in other ways) to survive, at least until the point where we have a stable society that doesn’t need to keep advancing to stay safe.
This set of arguments only establishes that undifferentiated progress is better than no progress. They do not by themselves directly argue against differential technological progress. However, people who ~roughly believe the above set of arguments will probably say that differential technological progress (or differential progress in general) is a good idea in theory but not realistic except for a handful of exceptions (like banning certain forms of gain-of-function research in virology). For example, they might point to Hayekian difficulties with central planning and argue that in many ways differential technological progress is even more difficult than the traditional issues that attempted central planners have faced.
On balance, I did not find their arguments overall convincing but it’s pretty subtle and I don’t currently think the case against “undifferentiated progress is exceptionally good” is a slam dunk[1], like I did a few years ago.
In absolute terms. I think there’s a stronger case that marginal work on preventing x-risk is much more valuable in expectation. Though that case is not extremely robust because of similar sign-error issues that plague attempted x-risk prevention work.
I think more important than the Hayekian difficulties of planning is that, to my understanding of how the term progress is normally used, part of the reason why all these notions of economic-technological, values, and institutional progress have been linked together under the term “progress” is because they are mutually self-reinforcing. Economic growth and peace and egalitarian, non-hierarchical values/psychology → democracy → economic growth and peace and egalitarian values, a-la Kant’s perpetual peace. You can’t easily make one go faster than the others because they all come together in a virtuous circle.
Further, these notions of progress are associated with a very particular vision of society put forward by liberals (the notion of long-run general progress, not just a specific type of moral progress from this particular law, is an idea mostly reserved to liberals). Progress, then, isn’t moving forward on humanity’s timeline, but moving forward on the desired liberal timeline of history, with a set of self-reinforcing factors (economic growth, democratization, liberal values) that lock-in the triumph of liberalism to eventually usher in the end of history.
I think this understanding makes the notion of undifferentiated progress much more compelling. It is hard to separate the different factors of progress, so differentiating is unreasonable, but progress is also a specific political vision that can be countered, not some pre-determined inevitable timeline. We live at a time where it feels uncontestable (Fukuyama a couple decades ago), but it isn’t quite.