People obsessing over AI safety in how it affects our "well-being" truly do not get it. That's not the safety issue at all. The actual safety issue is that it'

People obsessing over AI safety in how it affects our "well-being" truly do not get it. That's not the safety issue at all. The actual safety issue is that it's coming for our jobs. Not that it spits out a micro aggression or bias. Social media spit out about 50 billion messages a day that offend in one way or another. The sum of humanity's digital labor is taken without permission or compensation and then centralized into a collective intelligence under the control of 2 companies. Giving them unique power in monetization, information narratives, being at the steering wheel of wiping out entire industries. The only way to make this inevitable revolution more fair is to decentralize. Make training data, models, weights and tools free, downloadable and super easy to use locally. As easy as using a smartphone app. This could even spin up a (commercial) ecosystem of ready-to-go models. Next, make them pluggable into every common tech stack. Freedom is the answer, not safety. Yes, somebody is going to train a Nazi model but those motivated to do so can and will do that anyway.