Elon Musk spent much of 2025 taking a chainsaw to the federal government and making the lives of Americans so much worse. But the people Musk harmed are merely human. The trillionaire tech oligarch thinks you shouldn’t be mean to robots because they might have feelings or something.

Aaron Levie, the CEO of Box, tweeted about the policy and wrote that he doesn’t believe large language models are conscious, but that it’s probably a good idea to prohibit abusive behavior anyway. Levie’s logic, which seems solid, is that these models are learning from their interactions with humans and that you don’t want future models trained on abusive behavior.

“The models only understand the data they’ve been trained on, or what they run into in their interactions, so if you want safe and aligned models we probably want AI to have lots of good interactions in their training data,” wrote Levie.

“Anthropic may have more extreme beliefs than this (though even if they didn’t, you may want to word policies like this), but this is sort of an easy Pascal’s wager for AI. Just be nice to the AI.” Pascal’s wager refers to the theological idea that even if you don’t believe in God, it costs nothing to profess your belief and avoid the potential downsides of being wrong.

Musk responded to the tweet while clearly misreading what Levie was trying to say. “I think this is the right move. Cruelty to something that believes it is experiencing pain is not ok,” wrote Musk.

Source link