From the magazine

Real artists have nothing to fear from AI

Forget the supposed threat that AI poses to artists. The only question people should be asking is: can AI make good art?

Alexander Raubo
We can be superheroes: ‘Embedding Study 1 & 2’ (2024) by Holly Herndon and Mat Dryhurst Christie’s Images Ltd. 2025
EXPLORE THE ISSUE 01 March 2025
issue 01 March 2025

Christie’s is making digital-art history again – or at least trying to. Between 20 February and 5 March, it is hosting Augmented Intelligence, the first major auction dedicated solely to AI-generated art. This follows a series of headline-grabbing stunts, including the first major sale of an AI-generated artwork in 2018 – ‘Portrait of Edmond de Belamy’ ($432,500) by the Paris-based collective Obvious – and the first NFT sale by a major auction house,  Beeple’s ‘Everydays: The First 5,000 Days’, which shattered expectations (and good taste) by selling for $69 million in 2021. With the NFT bubble – which Christie’s played a significant role inflating – having burst in 2022, its attention is now turning to marketing another promising asset class, that of generative art. The question is: can AI make good art?

Generative art, of which AI art is a species, has been around longer than today’s NFT collectors might think. Rooted in 1960s conceptual and computer art, it remained a niche interest until the launch of the browser-based version of ChatGPT in 2022, which amassed 100 million users within two months to become the fastest-growing consumer application in history.

In classical generative art the artist would set up a series of rules and get someone to execute them – be it a computer, or a patient gallery employee. The charm came from the degree of randomness that inevitably ensued. Sol LeWitt’s ‘Wall Drawings’ had humans following orders like machines, producing subtly different results each time. Charles Csuri’s ‘Bspline Men’ (1966) applied complex maths to the drawing of a bearded man. Harold Cohen, meanwhile, created a program in the late 1960s called AARON and fed it increasingly complex commands related to how we see until it produced results such as ‘Untitled (i23-3758)’ (1987), a joyful, Cézanne-like composition.

GIF Image

Magazine articles are subscriber-only. Keep reading for just £1 a month

SUBSCRIBE TODAY
  • Free delivery of the magazine
  • Unlimited website and app access
  • Subscriber-only newsletters

Comments

Join the debate for just $5 for 3 months

Be part of the conversation with other Spectator readers by getting your first three months for $5.

Already a subscriber? Log in