Collective Action Problem

Isn't this problem of out of control AI rooted in the same capitalist contradiction as climate change? This dilemma of "alignment", trying to make sure the technology is in humanities best interest, is the same one we face with fossil energy. The Market, it shpould be obvious by now, is useless. And government is controlled by Capital. So a good liberal like Ezra Klein sees these labs competing with each other to achieve "recursive self-improvement", AI agents who take over their own training, and he calls for social "control". He asks why aren't their permits and regulations and governance working to serve the publics general interest? Well, why are we still burning fossil fuels: the collective action problem. With so much wealth and power at stake, no one actor is going to stop (building RSI, burning fossil fuel) unless everyone does. Kind of like a nuclear non-proliferation treaty. Give up your nukes and get crushed by the other nuclear powers (see Libya). Does Ezra Klein expect the United Nations to enforce a ban on AI development? He claims this development should face the same constraints as building a new high-rise in San Francisco. But this is apples and telephone poles: wholly separate categories. I would love it if these liberals would just admit market failure in both pollution and AI development and begin advocating for democratic planning. Tearing down these corporations and putting ownership of production in the hands of the people who make and consume. Then you would be able to exert genuine "control". You can't have it both ways. If you stick with Capital you face the growth imperative and therefore are doomed to competition and the collective action problem. The only way out is planning, not for profit, but for sufficiency. We all know perfectly well that infinite growth leads to doom but we act as if we don't know it. One day just leads to the next as we slowly crumble around the edges. Not unlike a melting glacier.