4 ms·
I get this was a design/art project and not necessarily meant to be sophisticated political philosophy, nonetheless there are a few distinctions here that would
by gampleman 4y ago
I get this was a design/art project and not necessarily meant to be sophisticated political philosophy, nonetheless there are a few distinctions here that would be useful:
To be ruled by AI I think requires the AI to be in some sense conscious, to have some notion of I and be able to form its own values and goals. Otherwise the ruler is whoever imposes/creates the AI, just in a once removed sort of way.
On the other hand to be ruled by algorithms: we already are. That's basically what the "rule of law" means - it's the (actually fairly desirable) right to be ruled and governed by a set of (in principle) knowable algorithms, which creates a fairer society than one ruled directly by human dictat.
What's being depicted (and criticised) in the film I think are mostly two things:
1. Some notion of social credit, where desirable actions are being rewarded with some arbitrary tokens and undesirable actions are being punished by these tokens being taken away. In some way this is already the case with capitalism (where desirable actions are work/investment and tokens are money), so I think some notion where what is societally desirable being more broadly evaluated isn't necessarily that terrible.
2. Persistent surveillance. This makes the above notions of social credit a bit more challenging since in some ways it makes participation in the social ranking system less voluntary (although how voluntary is work/making money in our own system is pretty debatable). Also it's creepy and dangerous in a number of other ways.
Neither of these things really require AI or algorithms to be bad, but AI tech really helps with the scalability of these solutions and therefore their applicability to real world societies. But it doesn't really seem like you are critiquing things for which AI would actually be bad as such (such as bias in training, value of compassion over justice, etc)
- bbor 4y agoGreat points, I think they used “AI” to be a bit provocative on purpose. IMO this is really more about taking the concepts of pure utilitarianism to their extreme, as enshrined by systems and technologies that rule our lives. This fits the term “artificial intelligence” very well I think, but in a very different way than how it’s used today (I.e. conscious AGI agents or stats-based ML models)
- hemmert 4y agoThanks for these points!
- _bkyr 4y ago> Some notion of social credit, where desirable actions are being rewarded with some arbitrary tokens and undesirable actions are being punished by these tokens being taken away. In some way this is already the case with capitalism (where desirable actions are work/investment and tokens are money) I would argue this isn't always the case. There's a lot of reward being given for things that are detrimental to society. Pharma companies being at the top of the list.
- ermir 4y agoRule of law differs substantially from rule by algorithms since laws are created and interpreted by humans, and there's always the possibility of reasoning and making a deal with a human when the law is incorrect or there are specific circumstances. In rule by algorithm, you get a situation where "Computer says no", and there's no recourse, no deal to be made, nobody to reason with. An entire society can be bootstrapped to be entirely run by thoughtless algorithms and it could end up in a situation where there's no way in to untangle the dependencies that these systems have with one another.
- gampleman 4y agoWhether and what recourse is available is pretty orthogonal to whether the thing is run by computers or humans. In practice it is likely to be a mixture of both these days, as we don’t really have sufficient technology to run a real society, and most countries today take heavy advantage of automation.