Machine Hallucinations- MRO Dreams – A by Refik Anadol
Trends come and go in the art world, and it can be somewhat difficult to predict which ones will stay around and which will fizzle out. Because of that, I’ve found it useful to have skepticism as my first initial reaction. Some recent contemporary art sales featuring NFTs and AI-generated art have served as an interesting example of how quickly trends in the secondary market can come and go.
Last week, on February 28th, Sotheby’s New York hosted their Contemporary Discoveries sale, featuring works by artists like George Condo and Wolf Kahn. However, nineteen of the two hundred sixty-five available works were NFTs. With the world of blockchain technology suffering many setbacks over the last few years, NFTs are no longer the hot new darling of the art world. That can be seen in the results of just these nineteen works. Sotheby’s specialists predicted these NFTs would bring in at least $256K hammer, yet they made just $80.7K (or $102.5K w/p). These shortcomings can mainly be attributed to the failure of Refik Anadol’s work Machine Hallucinations- MRO Dreams – A, which Sotheby’s gave an estimate range of $120K to $180K. Even with the Anadol removed from the equation, it’s still not a great look for NFTs. Sixteen of the nineteen NFTs sold. However, thirteen sold below their estimates, sometimes incredibly far below them. Works valued at a few thousand dollars were not generating the expected enthusiasm from buyers. Works estimated to sell for at least $5K were selling for $2K, or even as low as $400 in one case. The failure of these NFTs was part of why the sale overall proved dissatisfactory for Sotheby’s, with 38% of the lots selling below their estimates.
I bring this up because Christie’s, despite an incredible amount of pushback, decided to go ahead with their AI art sale on Wednesday, March 5th. Several publications seem to have declared the sale an unequivocal success when it’s actually a bit more complicated than that. The sale had an 82% sell-through rate, which is rather good. However, fourteen of the thirty-four available lots sold for below their estimates, amounting to 41% of the entire sale. Furthermore, there seems to be some confusion about the sale’s total. When I first wrote about the sale while writing about the artists’ petition calling for Christie’s to cancel the auction, I cited a number I had seen in several other publications, saying that the sale was expected to make around $600K. However, after totaling up all thirty-four lots, that number is actually $806.1K. While that isn’t an insane difference as far as major auction house sales go, it does make a difference here. The total hammer price fell short of this estimate, adding up to only $578.4K. However, because the sale’s total with all fees added brought that number up to $728.8K, many covering the sale see this all as a success since that number is definitely more than $600K. But I’m unsure where this $600K number came from. I haven’t been able to find the source of this. It may have been an approximation before any additional lots were added to the sale that would augment that number.
So, the AI art sale was not as successful as people thought, yet many are treating it like it is. Because AI art is now the new hot topic, it’s the latest trend. And like I said before, it’s always safe to be skeptical. Artificial intelligence may be impacting our world in various ways, but I don’t think that collectors overall will ever be willing to fully embrace AI-generated art to the degree that some are predicting.
AI Art: The New NFTs?
Machine Hallucinations- MRO Dreams – A by Refik Anadol
Trends come and go in the art world, and it can be somewhat difficult to predict which ones will stay around and which will fizzle out. Because of that, I’ve found it useful to have skepticism as my first initial reaction. Some recent contemporary art sales featuring NFTs and AI-generated art have served as an interesting example of how quickly trends in the secondary market can come and go.
Last week, on February 28th, Sotheby’s New York hosted their Contemporary Discoveries sale, featuring works by artists like George Condo and Wolf Kahn. However, nineteen of the two hundred sixty-five available works were NFTs. With the world of blockchain technology suffering many setbacks over the last few years, NFTs are no longer the hot new darling of the art world. That can be seen in the results of just these nineteen works. Sotheby’s specialists predicted these NFTs would bring in at least $256K hammer, yet they made just $80.7K (or $102.5K w/p). These shortcomings can mainly be attributed to the failure of Refik Anadol’s work Machine Hallucinations- MRO Dreams – A, which Sotheby’s gave an estimate range of $120K to $180K. Even with the Anadol removed from the equation, it’s still not a great look for NFTs. Sixteen of the nineteen NFTs sold. However, thirteen sold below their estimates, sometimes incredibly far below them. Works valued at a few thousand dollars were not generating the expected enthusiasm from buyers. Works estimated to sell for at least $5K were selling for $2K, or even as low as $400 in one case. The failure of these NFTs was part of why the sale overall proved dissatisfactory for Sotheby’s, with 38% of the lots selling below their estimates.
I bring this up because Christie’s, despite an incredible amount of pushback, decided to go ahead with their AI art sale on Wednesday, March 5th. Several publications seem to have declared the sale an unequivocal success when it’s actually a bit more complicated than that. The sale had an 82% sell-through rate, which is rather good. However, fourteen of the thirty-four available lots sold for below their estimates, amounting to 41% of the entire sale. Furthermore, there seems to be some confusion about the sale’s total. When I first wrote about the sale while writing about the artists’ petition calling for Christie’s to cancel the auction, I cited a number I had seen in several other publications, saying that the sale was expected to make around $600K. However, after totaling up all thirty-four lots, that number is actually $806.1K. While that isn’t an insane difference as far as major auction house sales go, it does make a difference here. The total hammer price fell short of this estimate, adding up to only $578.4K. However, because the sale’s total with all fees added brought that number up to $728.8K, many covering the sale see this all as a success since that number is definitely more than $600K. But I’m unsure where this $600K number came from. I haven’t been able to find the source of this. It may have been an approximation before any additional lots were added to the sale that would augment that number.
So, the AI art sale was not as successful as people thought, yet many are treating it like it is. Because AI art is now the new hot topic, it’s the latest trend. And like I said before, it’s always safe to be skeptical. Artificial intelligence may be impacting our world in various ways, but I don’t think that collectors overall will ever be willing to fully embrace AI-generated art to the degree that some are predicting.