Firstly, “…remotely resembles…”. is clearly an exaggeration. Under what test conditions? How would we know without double-blind testing with a sufficiently large audience. Digital and wet prints can resemble each other very closely, although there are some giveaways. Some digital prints exhibit changes in reflectivity in different places, and as I’ve mentioned before, wet prints are seldom flat.
Is this what concerns us? Should we spend our lives squinting sideways at prints to decide on their merits? How much does it matter?
I’m inclined to agree with Joanna that we have reached equality. By this, I mean equality between competent practitioners.
One note - I am speaking here specifically and only of monochrome film projected onto silver prints. Nothing more.\
No paper on which I have seen digital media reproduced comes close to the look of a silver fiber print. I stipulate I have not seen all- or most candidates, but enough to be convinced looking further isn't worth the effort.
When I say "look" I specifically mean how the silver print renders the tonal range, how it reflects light, and how it hangs on the wall. It may, in fact, be inferior by some clinical definition to digitally printed media, but I just don't like the look.
That's not to say a digitally produced physical print cannot be beautiful - it can be - I just find it far less satisfying than the great silver prints I have seen - at least for that set of digital media I have sampled. No matter how hard they try, metamerism seems to show up to interfere with the work.
I saw the a show of Vivian Maier's work. The stuff done chemically was astonishingly good. Right next to it was stuff that had been produced from scan and digital printer. It was good, but clearly inferior, to my eye at least.
The flip side is something digitally captured and then projected onto silver paper. Here I have a somewhat different view. The final thing does look like a silver print. But - again noting I have not seen everything possible - the digital capture itself looks so different than film response, I actually find it qualitatively looks more like a giant Ektachrome image than something from a monochrome silver negative. The dynamic range, the linearity of the capture, and the edge effect are - to my eye - very 'chrome like.
Let us think of AI. It’s very early days indeed for AI. It will get better. Already it’s better at diagnosing diseases than doctors.
Let’s suppose we encounter an object on a gallery wall. It’s described as a Work of Art. You like it very much. You inspect it closely. The gallery owner introduces you to the person responsible for making it. As you shake hands, you notice traces of paint under their fingernails.
Given everything you have seen, is that person a painter or a computer programmer who does DIY? Does this alter the object on the wall?
Creation by mechanical means is not a new idea. Julia, in 1984 (pub 1949), operated a machine for writing novels.
But that's not the question. When mechanical means are controlled by an artist, then it's manmade art. But the application the AIs as currently envisioned is headed in the direction of completely hands of interpretation of the scene. I stipulate that you may get something beautiful out of that but I'd argue it's not "art". Art, definitionally, must come from the hands of mankind. Otherwise monkey feces would be hung at MOMA ... oh ... wait.
The critical question isn't whether they wrote code or used a brush - both are legitimate ways to make art. The critical question is
whether they were in control of the process which is decidedly not the case with AIs even if they
did author the large language model or rulesets. There is a kind of deep question about whether AI even qualifies - strictly speaking - as an "algorithm'. An algorithm, again by definition, must produce the same outputs for a given set of inputs. It's arguable that AIs do not do this. (That's a little pedantic, I know, but it speaks to the question of who is controlling the "art".)
