As you may have seen in the news, last week a painting that was (sort of) created using artificial intelligence (“AI”) sold at Christie’s in New York for a whopping $432,500.
Christie’s take on the sale is rather disturbing, but then of course, they have a vested interest in what they’ve done here, and of course in the fee that they’re about to collect for their efforts in this charade. “It may not have been painted by a man in a powdered wig,” commented Christie’s head of prints and multiples Richard Lloyd, “but it is exactly the kind of artwork we have been selling for 250 years.”
Well no, it isn’t.
Let’s assume that this was in fact an Old Master painting, created by an actual French artist working in around 1760, and portraying a real French aristocrat. It’s a blobby mess of an image, with no particular distinction in terms of technique or composition, and the top of the picture appears to be cut off. Is Christie’s really maintaining that it would have gone to all of the effort that it has to market this work back in the 18th century, if it was just some chopped-off, undistinguished oil study by an unknown 18th century artist of an obscure, minor member of the French nobility? Such a claim is utterly lacking in plausibility.
British art historian Bendor Grosvenor who, despite my occasional differences with him, is still the best online read when it comes to looking at Old Master paintings in the contemporary world, is all over this story, as you might imagine. As he points out here, not only is the “AI” aspect of how the image was created somewhat suspect, but the portrait itself is little more than a Photoshop project:
But the much vaunted ‘AI’ artwork at Christie’s, Portrait of Edmond Belamy, is little more than a composite blurring of the 15,000 portraits fed into the programme in the first place. It’s you or I fiddling around on Photoshop for an hour, just scaled up. A regular cry against much contemporary art is ‘my child could have done that’. But now we can replace that with; ‘my laptop could have done that’.
Over on ArtNet, Tim Schneider describes Christie’s actions as “reactionary”: not in the sense often used by the left to describe conservatives, but rather in the sense of someone jumping on the first available bandwagon, as it were. “To me, this is about as reactionary as looking up from a fortune cookie promising true love and proposing to the first person you see in Panda Express,” Schneider scoffs, “and it speaks volumes about how superficially the high-end of the market is engaging with art and machine learning right now.”
I won’t get into the issue, pointed out by Grosvenor and Schneider, among others, that a significant segment of the art community is bemoaning the fact that the “first” AI painting to be sold at a high-end auction features the image of a (supposedly) dead white man. For those of you who are interested in such things, there are plenty of comments to that effect scattered across social media. Yet wherever you fall along the social justice warrior spectrum, I don’t think that issues of race or gender are really the point here.
Rather, in this story we have further proof, as if that were needed, that the Contemporary Art world is first and foremost a speculative bubble. It continues to be inflated by auction houses, art dealers, art media, and the art establishment. The goal is not to celebrate and encourage the creation of great works of art, but rather to make money off of poorly-educated, socially insecure, extremely wealthy people.
The idea of the art world bilking the newly rich by convincing them to purchase art at inflated prices has been around for a long time. Renowned art historian Bernard Berenson (1865-1959) was giving intentionally iffy or flat-out wrong attributions on works of Italian Renaissance art to Sir Joseph Duveen (1869-1939), art dealer to the robber barons, a century ago. That way, Duveen could jack up his prices, and Berenson could receive a larger fee for his services. Colin Simpson’s superb book, “Artful Partners” is still the authoritative text on this nefarious arrangement, which was the most significant, but not the only, one of these sorts of arrangements that existed during the Gilded Age. The Faustian bargain between the two deceivers ended up having a negative impact on art history for decades, and their actions still carry repercussions for art scholarship even today.
Recently I was fortunate enough to acquire the sort of painting that Duveen might have carried in his gallery a century ago. It’s an oil on wood panel depicting St. Jerome in the wilderness, engaging in prayer and penitence, and accompanied by his iconic lion. It dates to somewhere toward the end of the Italian Renaissance, and although reminiscent of the work of artists such as Palma il Giovane (1548-1628), I’m by no means expert enough to make such a firm attribution. It was purchased by a wealthy Pittsburgh manufacturer from a European art dealer about a century ago, and donated to a Rust Belt art museum, which has now deaccessioned it for a hammer price that was no doubt considerably less than what the American collector who brought it across the Atlantic originally paid for it, in today’s dollars.
If our art collecting Northeastern industrialist were alive today, and of course he would not be in heavy industry but in some sort of digital business, the AI portrait sold at Christie’s last week would be exactly the sort of thing that an art advisor would recommend that he purchase. The impetus to acquire however would not be because the art in question was actually any good, but because his owning it would attract the notice of the public and the admiration of his peers. There is no such thing as “bad” publicity, in the present age.
There is a difference here, however, between the world of the early 20th century buyer and that of the early 21st century buyer. While a beautiful image of sacred art is always going to find an audience, so long as there are still Christians who treasure such things, it would surprise me to learn that someone living a century from now would pay the equivalent of nearly half a million dollars for an unremarkable piece of computer art. Perhaps I am wrong about that, of course, but fortunately, I won’t be around to find out.