Garrison Lovely argues that opposition to AI should focus less on shaming individuals for using it and more on organising against companies pursuing systems designed to replace human labour. He says the industry’s stated ambition to build artificial general intelligence amounts to creating a machine capable of doing almost any work that can be done remotely, putting many livelihoods at risk.

Lovely calls for a freeze on frontier AI development until there is strong public support and broad scientific agreement that it can be made safe and controllable. He argues that refusing to engage with AI leaves its use and governance to industry leaders and others who do engage: people can use the tools without endorsing the companies behind them, and can better understand their capabilities and influence policy. He cites public comments by AI executives and policymakers, and notes that 69 serving members of Congress have publicly referred to concepts such as superintelligence, the Singularity or AI existential risk.