Is it ethical to work on general-purpose robots given the risk of cyberhacking?
One potential risk of developing general-purpose robots is that they could greatly reduce the friction required to establish a totalitarian regime. If there were millions+ of general-purpose robots deployed in the world (in households, nursing homes, construction sites, etc.), and someone were able to hack / take control of a large fraction of them, they could easily create a repressive state through telling the robots to oppress humans (e.g., arm themselves with weapons and threaten humans).
(To clarify, I do not mean “general-purpose robots” as a synonym for ASI. I define “general-purpose robots” as robots capable of learning human-like physical tasks on the fly, such as being placed on a worksite and quickly learning new construction tasks. This ability to quickly learn new tasks could make them dangerous if they were compromised — for example, a hacker might teach such robots to pick up weapons and intimidate people.)
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 the cyberhacking of general-purpose robots is considered a serious risk—and if so, whether it is ethical to work on general-purpose robotics; or 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 deeply appreciated.
I think you’re worrying too much about a tiny problem.
There’s a huge industry vigorously working on killer robots and robotic tools for oppression. If you actually do robotics, you will soon meet them, and have fun chats at conferences. The threat of robots being subverted by hackers is so minor in comparison that you can ignore it.
I worked in and around robotics for a number of decades. The position I eventually came to was that working on robots for commercial or artistic purposes was Good but working on robots for the police or military was Bad. A commercial robot could be repurposed to be military, but if someone did that, it wasn’t my moral culpability. The same applies to developing general robotic algorithms like SLAM. As long as they can be used for purposes that are obviously good, that suffices to feel good about developing them.
Maybe I’m wrong and you’re right, and the risk of robot subversion by hackers is substantial. In that case, you can do much good for the world by working in robot-specific cybersecurity. It’s not a reason not to do robotics in general.