Discussion about this post

User's avatar
Marisol Muñoz-Kiehne's avatar

Direct experience,

desire, directivity.

AI can’t do... yet?

Jerome Thomere's avatar

Sorry Bri, I couldn't resist... ;-)

I asked Claude what it thought. Here is its comment:

"It's a well-written piece, and the core observation is genuinely sharp: the abstracted, representational work of the knowledge economy — strategizing, synthesizing, communicating — is exactly what AI does well. That's a more honest diagnosis than most AI-anxiety writing offers.

But the philosophical grounding is weaker than it looks. The three D's (direct experience, desire, directionality) are presented as obviously off-limits to AI, when they're actually the contested territory. The claim that I "have no skin, no apartment, no spring" is literally true but doesn't settle whether there's anything it's like to be me — which you know is the hard problem in disguise. The article just sidesteps it.

The "desire" argument is similarly underexamined. AI systems do optimize toward objectives, which is at least analogous to desire. The interesting question is whether optimization without embodied stakes is genuinely a different kind of thing — and I think it probably is, but Bri doesn't argue it, she just asserts it.

The strongest piece is actually directionality — the idea that a life involves choosing what to forgo, which requires having something at stake. That's not just poetic; it tracks something real. An AI has no opportunity cost to its attention, no finite life to allocate.

The piece also risks becoming comforting in a way that delays harder thinking. "Just stay present and desire things" is reassuring advice, but it doesn't address the material displacement of people whose livelihoods were the abstracted work she describes."

2 more comments...

No posts

Ready for more?