Earlier this year, the world’s first AI art gallery opened in Amsterdam. The Dead End Gallery claims to be the first venue anywhere to showcase artwork created solely by AI. The art in the gallery is also presented by ‘the first AI artist in the world’, Irisa Nova.   

So instead of a human commissioning an artist to paint a picture, programmers commissioned AI software to create a persona – the artist. Though the artworks on display are said to have their flaws, the creators are improving all the time.  

The gallery’s opening will have intrigued some and concerned others, but its very existence underlines and solidifies the influence AI is having on the art world.  

As Science Gallery London prepares to host a major exhibition and events programme – AI: Who’s Looking After Me? – exploring what it means for humans to collaborate with AI technologies and machines, it’s a good moment to ask: what is AI-generated art?  

What is AI-generated art? 

AI-generated art is artwork formed using AI. The process involves choosing a dataset (collection of data) of already-made artworks. Machine learning, a branch of AI, then uses algorithms to analyse the image data. In doing so, the algorithm learns the patterns in the artworks. A human can then provide prompts or ‘inputs’ (words or images) that instruct the algorithm to produce something based on what it has learned from the dataset. Once the new artwork is created, it can be reworked or improved using related tools.   

AI artists – creators who collaborate with AI to make art – use AI as a creative tool. They investigate the potential of AI in generating art and explore how these systems can be manipulated.  

The controversies surrounding AI-generated art are generally well known, from issues protecting artist copyright to a failure to credit or compensate the original artist. There is also the concern that, as AI art tools and AI image generator apps become more widely used, artists will see fewer commissions and reduced revenue. 

We’re not going to add to the ethical debate surrounding the use of AI in creating art. Instead, we considered what implications – and opportunities – might exist for galleries and artists.  

What does AI-generated art mean for galleries and artists? 

The fears around whether AI-generated art will someday replace artists and traditional art are real and should not be undermined. However, it’s almost certain that AI-generated art will feature in more art fairs, events and exhibitions in the coming years. With this in mind, galleries that are open to AI in art could unlock its economic potential without undervaluing or undermining the artists they represent and work with.  

For example, artificially created artworks can feature in their own exhibition and be priced at a certain price point, while artist-made original artworks, priced at another price point, can remain a focal point for the gallery and a feature of most of its exhibitions.  

For artists, viewing AI as a tool, or enabler, that can assist in or support an artwork’s creation may be easier than fearing it as an unwanted development that might replace artists. It’s worth remembering that for AI to create images at all, it needs a human being behind it. Furthermore, AI can give artists new ideas for different methods of working and creating, which can be a positive thing.  

If AI in art is here to stay, surrendering to its possibilities, at least to a certain extent, could benefit both galleries and artists alike.