This is neat but I don’t think it supports the argument they claim they make in the paper. But also their links to the paper are broken so I can’t check.

No one would say this 1 bit circuit has “human” attributes regardless of substrate, I think it’s a matter of scale.

Because for some reason they didn’t even link their paper on their own website: https://arxiv.org/abs/2605.31514

At this point people are having semi-theological arguments about whether the LLM machine spirits can exist in every Turing complete system, I guess.

Computer science sure has gotten weird

You’d think if anything they’d make the argument with something far more absurd, like that computer made out of live crabs. “Do crabs start being human-like if we make them move in the correct ways” would be a more fun argument to have