What Then Must We Do About AI?

The UK creative industries are financial powerhouses, recognised internationally for writing, art, video games, film, television and other creative talent. For proof, you just need to look at the numbers. Government statistics show that the creative industries grew by 6.8% in 2022 and contributed £124.6bn to the national economy. That’s more than the UK car industry, or aerospace, oil and gas.The landscape for creators themselves, however, is not so bright.  

Socioeconomic factors such as the decline in cinema and television audiences, the impact of COVID-19, streaming models, social media advertising, and ironically, the US Actors and Writers strikes, are squeezing employment to the point where 75% of the UK freelance media workforce is out of work. The US Art Directors Guild has even paused its training courses, stating that it ‘cannot in good conscience encourage you to pursue our profession while so many of our members remain unemployed.’ Amongst all that, there was a global pandemic too. 

And then, generative artificial intelligence appeared. 

The People Behind the Code 

My journey with generative AI began in 2019, while I was conducting research for a South West Creative Technology Network (SWCTN) Fellowship. I was investigating automation within the creative industries, with a specific focus on AI and origins of the data Google was using to train its systems. At the time, it felt like big tech companies were sneaking behind the counter and stealing all the sweets. I shared these concerns in Who Owns Digital Stories (2019) and co-wrote a white paper on Unconscious Bias in AI Data Sets, and proposed a sociotechnical approach to managing its inevitable appearance. 

Much has changed over the past five years, and yet in many ways, nothing has changed. Authors and artists are increasingly aware that large tech companies are using copyright materials to train their AI systems, often resulting in text and images that are remarkably similar to their own. But the legal precedent has already been set. In 2016, a ruling determined that while AI systems were absorbing copyrighted material, what they were doing with it was “transformative” and therefore not a violation of copyright law. In my non-legal opinion, this ruling marked the end of any subsequent legal challenges about training data against big tech companies. 

Another significant shift is AI’s transition from a game of innovation to one of political power. It is a technology capable of influencing elections, advancing nations, accelerating productivity, and making certain individuals very rich. It is simply too big to fail.  

A common question is whether it will take our jobs. The optimistic response is: “AI won’t take your jobs, but someone who knows how to use it will.” The reality is: “AI won’t take your jobs, but the people who make it will.” I emphasise the role of the people behind AI because the technology itself is not inherently harmful. AI is just computer code, without a moral compass. However, the people who run the companies, and the programmers who write the code, do have biases, and their code does precisely what they tell it to do. 

These biases become apparent when AI generates images of racially diverse Nazis or exclusively white businessmen. Such results reflect the biases of the predominantly white, male engineers in California who write the code, funded by white male venture capitalists, who use terms like ‘moonshot’, ‘kill the competition’ and ‘world domination.’ Anti-creative biases come from leaders like Mira Murati, the former Chief Technology Officer (CTO) of Open AI who recently said, ‘Some creative jobs maybe will go away, but maybe they shouldn’t have been there in the first place.’ Similarly, Eric Schmidt the ex-CEO of Google once remarked, ‘If nobody uses your product it doesn’t matter if you stole all the content. Don’t quote me on that.’ 

Creating an Alternative Future 

As Tolstoy asked in his short book on society and greed, What Then Must We Do? 

For me, digital ethics and philosophy must be at the core of every project. This can be simplified by the famous Jurassic Park quote: ‘Scientists were so preoccupied with whether or not they could, that they didn’t stop to think if they should.’ 

The pace of innovation in the digital world is faster than I have ever seen in my 30 years in the industry. It’s exciting, exhausting, motivating, and overwhelming all at once. From my SWCTN research, my answer to Tolstoy’s question was simple: “Something!” 

My conclusion was that merely criticising AI developments or blocking them was not the solution. A different narrative was needed. We built on our 2019 paper, AI Ethics in Data Sets, and experimented with generative AI through our art application, Project Electric Sheep. We also explored interactive storytelling using our immersive video game adaptation of John Wyndham’s classic sci-fi novel, The Kraken Wakes, and developed Charisma.ai as an interactive AI-supported writing platform. 

Then, in 2022, ChatGPT launched, and everything changed once again. 

 With ChatGPT, the power of generative AI became clear to everyone, and questions about how OpenAI gathered its data gained prominence. This was the inspiration to create “something” – an active response rooted in the values of the creative industries while harnessing the capabilities of generative AI. Our vision was not merely to create stories, but to create new forms of storytelling and generate new revenue streams for creators. We aimed to bring these stories to life through words, visuals, and soundscapes. 

This month, we announced R&D funding from Innovate UK to take the first steps toward this vision. We are achieving this through workshops with creators of all ages and backgrounds, partnering with companies such as Aardman Animations, Channel 4, and Sound Reactions, as well as research collaborations with Falmouth University’s Film & TV unit and the Creative Computing Institute at UAL. Importantly, we also ensure a digital ethicist, Lisa Talia Moretti, is present throughout, overseeing our ideas, decision-making, and impact. 

Mapping the Future of Creativity 

The Charismatic.ai project aims to provide answers for how the creative industries can evolve in a world where AI plays an integral part in the creative process. Some of the answers will be joyful and exciting, ushering in a new era of creativity. Others may present challenges. However, through Charismatic.ai, we will create a map to help navigate this next phase of creative thinking. 

The key insight I took from my SWCTN Fellowship was the relationship between cause and effect. AI is the cause, and automation is the effect. While the sheer volume of information about AI can be overwhelming, studying automation offers a valuable perspective on how we can improve our processes and create work that would not have been possible without it. 

The most exciting part of this journey is this: how can we tell new stories with this new technology, and become better storytellers along the way, exploring how the ‘industry’ in creative industries can be reformed and strengthened to favour creators.  

That is something worth fighting for!