Hmm. I don't want to boycott AI products in general, because they can be incredibly useful, which makes it kind of hard to get people to boycott them.
Boycotting certain classes of AI products or services, such as replacing humans in certain tasks might be more realistic in the sense of potential for getting sufficiently large amounts of people joining the boycott to put pressure on companies and maybe affecting societal and policy changes. Something along these lines might be for example to encourage a social norm to always demand human review of even the tiniest bit of automated decision making (GDPR mandates a limited right for individuals to not be subjected to purely automated decision making.)
A related thing would be labor organization actions such as strikes, which could be used to protest use of AI, but also in that, I think that replacement of very large fractions of human labor in some sectors is unavoidable even if nothing stronger than gpt4o is ever trained, because there are such enormous advantages that can be had in some cases, and because capabilities that gpt4o the chatbot etc don't have, can be achieved with various strategies, such as combining them with all kinds of plug-ins, which can also include other, specialized llms or other neural nets. In a lot of cases, this would not mean replacing all human workers, but perhaps a small team could supervise a bunch of robots etc, and maybe some of them would have to get their hands dirty some of the time. Which sectors this will hit, and to which degree, is hard to predict. Also, it might create some new jobs, but probably more would be lost.
Because of these factors I think that whatever power can be wielded through boycotts or labor actions, should not be wasted in a futile fight to ban replacing humans.
Instead... To be continued
