Is it ethical to work on general-purpose robots given the risk of totalitarianism?
One potential risk of developing general-purpose robots is that they could greatly reduce the friction required to establish a totalitarian regime. If robots became physically capable of manufacturing additional copies of themselves, a small group of bad actors could potentially manufacture millions of general-purpose robots and use them to establish a repressive state — for example, by arming them and using them to coerce the population.
(To clarify, I do not mean “general-purpose robots” as a synonym for ASI. I define “general-purpose robots” as robots capable of i) learning human-like physical tasks on the fly, such as being placed on a worksite and quickly learning new construction tasks and ii) manufacturing more copies of themselves when instructed. They would not have to self-replicate autonomously; it would be enough for them to have the physical dexterity and manufacturing capabilities to build additional robots when directed by humans.)
When I tried to find assessments of this risk pathway online, however, I found very little; there is some discussion of the ethics of developing autonomous lethal weapons, but none on the ethics of developing general-purpose robots more broadly.
For this reason, does anyone happen to know: i) what the mainstream AI safety community’s view on this issue is — i.e., whether it is ethical to work on general-purpose robotics, given its ability to increase the baseline risk of sliding into totalitarianism; and ii) any good online discussions that examine this specific issue?More generally, I feel like there has been relatively little discussion in the AI safety community about the ethics of working on robotics, so any relevant resources would be appreciated.
(Note: Made a similar post previously about risks involving cyberhacking of general-purpose robots. In this post, however, I am asking specifically about the risks of general-purpose robots in general.)
Can’t speak for the “mainstream AI safety community”, but I personally think it’s bad to work on general purpose robots.
“Risk of totalitarianism” is even underselling the danger somewhat. General purpose robots will 1) help oppress and defeat people directly 2) substitute for soldiers, thus removing one of the pillars of human irreplaceability and political power 3) substitute for workers, thus removing the other pillar of human irreplaceability and political power.
And I wouldn’t focus on “a small group of bad actors” either. The discourse has been so poisoned by this “rogues rogues rogues” fearmongering, with a side order of “China”. In reality, the most likely group to use robots this way is simply today’s rich and powerful elites, who would love to have no need for (and no fear of) the rest of us.
I think I was able to briefly see the picture for a few minutes some months ago
I’d say there’s been some discussion, for example these, all from 2025:
acute: AI-enabled coups, Davidson et al
gradual: Gradual Disempowerment, Kulveit and Douglas et al
possibly: Law-Following AI, O’Keefe et al (discussion of ‘henchmen’, I can’t remember if they go into robots)
(and various discussions following those, as well as earlier discussions which are less salient in my mind right now).
Gist is basically what cousin_it says: oof, scary, because it could directly surveil+oppress, and because it erodes widespread human leverage-via-usefulness (as workers or soldiers), and because in replacement contexts it could throw whistleblowing and qualms out the window with the displaced human (these are all interlinked).
General-purpose isn’t necessarily the crucial matter, but obviously makes all of the above more straightforward and obvious.
Generalisable argument against almost all forms of “super-technology” (e.g. extreme force multiplier style tech):
1. the world mostly works as it is; if it remained broadly similar to today, or perhaps say some year just before it seemed like we might “malthusian disaster” the world, we could potentially tick along fine forever, main risks being things like pandemics or pre-existing force multiplier stuff like nukes
2. extreme force multiplier tech (including now-mundane things like say, mass car travel, powerful industrialisation) has nonzero risk, in some% of universes it ends the world
From that—well, if you have a button that makes people x% happier in 95% of cases, but in 5% of cases untold horror occurs, the entire universe simply ceases to exist, etc, do you press it? What value of x do you need?
The main counterargument I think is that we already have a risk of total destruction (e.g. climate change) and so the “do nothing” path is not riskless.