in ,

Is artificial intelligence the future of the arts?


Professor Drew Hemment (picture: FutureEverything) and Dr Matjaz Vidmar (picture: Julija Pustovrh)

Professor Drew Hemment (picture: FutureEverything) and Dr Matjaz Vidmar (picture: Julija Pustovrh)

Professor Drew Hemment is Project Director and Principal Investigator of The New Real and Dr Matjaz Vidmar is Deputy Director and The New Real Co-Investigator

WELCOME to the latest generation of artificial intelligence (AI) technology shaking up the arts world. Image generators such as DALL-E and Midjourney are powerful tools that through a combination of data mining and analysis link ideas with visuals. By typing a few words anyone can generate high-quality images. Across the world, this is causing shockwaves and widespread fears of the loss of jobs for human artists. Have we indeed reached a watershed moment in the evolution of AI Art? Yes, but it’s not the one you think.

The current tools are incredibly powerful, but narrow in scope. To generate further artistic interpolations, they are dependent on datasets of human-made images for style and content. So, while there are genuine concerns about some jobs being lost to automation – specifically in commissioned illustration, where AI can quickly follow very detailed briefs – there are also exciting opportunities for creative collaboration between humans and machines.

Music streaming service Spotify is building new AI-powered tools to allow users to personalise their listening experience; in games such as Black&White players can interact with an AI avatar whose behaviour changes in response to how it is being treated. At the same time, artist Holly Herndon has challenged the boundaries of “traditional” digital music by intimately involving algorithms in the production of her music, while Jake Elwes has explored how spectra of non-traditional identity intersect with image generation.

These works not only extend the reach of AI, but they also examine real issues about AI amplifying bias and inequalities in society today, about ownership of technologies by a small number of multinational corporations, and about the authorship and rights of artists whose data is used to train these ‘automated’ systems. Artists such as Hendon and Elwes are at the forefront of efforts to engage audiences with these topics and to highlight and tackle these challenges.

In The New Real, a unique hub for AI and arts research in partnership between the University of Edinburgh and the Alan Turing Institute, we are working to imagine and deliver the next generation of creative AI tools. These tools are designed collaboratively between data scientists and artists to fuel transformative experiences for audiences, and also to develop technologies that are more inclusive and fair. We bring all this together at festivals – in Edinburgh and around the world – and present bundles of new technologies, new creative works and new kinds of audience participation.

We are here to offer platforms to open up the technical and artistic development to more people – though designing new AI tools and collaboration methods. This is also important outside the arts, as art is often the stepping stone to understanding the wider societal context of new ideas and technologies. One of our significant themes is sustainability – how can AI consume less energy and other resources, whilst still delight and engage audiences worldwide.

After all, it may be that the cost of electricity directs the scale of future development.



Source link

What do you think?

DALL-E raises tough questions for art and artificial intelligence

Viral chef raises over 3k to start London’s first TikTok restaurant