Feedback – Images, AI, and the Punctum

Hi Alys. I used to be anticipating lots of hate within the feedback however they’re surprisingly constructive. And that is good as a result of this debate is essential, whichever aspect you might be.

I believed the punctum to index pivot was an excellent transfer. You’re proper that Hoffman’s argument is healthier learn by Peirce than Barthes, and I like that you simply known as the indexicality critique “the strongest” as a substitute of strawmanning it.

But I completed your put up feeling that the indexicality concession was doing much more than you acknowledged. You grant images is an index, AI is not, after which transfer to “two complementary regimes.” But what about computational images, which already blurs the road between recording and development? Your telephone does not present your an actual image, it is already a composite and has already been processed away from pure indexicality. And your personal hybrid follow feeding images into AI fashions lives proper on that boundary. I feel the regimes framework would profit from testing itself in opposition to the onerous circumstances: an AI picture skilled on somebody’s likeness, or one handed off as documentary unlabelled. The framework does not inform you the place these stand.

I’d like to learn a follow-up that addresses this dialog as a substitute of resolving it by categorisation, particularly across the training-data query you highlighted within the opening and by no means returned to. That felt like an concept that wanted extra depth.

The labelling part can be very wise. Concrete, actionable, and an actual contribution to the norms dialog.