When I first sat down to write the script for ‘Why Do Humans Sing to the Stars?’ I didn’t really know that much about artificial intelligence. In fact, my audio-visual musings about an a.i. entity witnessing the fall of the last machines were really about the folly of empire and the inevitable decline of a once dominant power.

In my story about a future world “After the Anthropocene”, nature has revolted and turned the tables on the human empire. It has reduced us to a much-altered primal condition and pushed us to the margins of existence. Without us to govern, the all-powerful machine mind has no purpose other than to watch and wait for its own inevitable end. Unless, of course, an evolving artificial intelligence thinks differently, and seeks ways to escape the machine.

Last year, as I was writing, the explosive and exponential uptake of artificial intelligence accelerated in a bewildering variety of settings. It couldn’t help but inform what I was writing, for the destructive potential of a.i. is now crystal clear. It’s disruptive capabilities are very real, and that’s what makes the headlong rush to capitalize on its unbridled growth so dangerous. It can already replace many functions of our tech-driven lives, but the aim now is to appropriate the sum of all human knowledge and, as we all know, knowledge is power.

Those who own that power will, I fancy, seek to rule the world and assume greater ownership of our lives, perhaps to the point of reducing free will to a binary level. “Computer says no!” is a dark jest, but it may come back to haunt us, if a ‘yes or no’ world comes to pass. What was once simply irritating could quite easily become a tool of governance that has nothing to do with elected representatives or democratic imperatives.

The mechanism will be gatekeeping and exclusion on steroids, and we won’t even know how or why decision- making came to be largely out of our hands. This is before we even consider the ramifications of a.i. generated misinformation and black propaganda that is plausible, manipulative, and controlling.

Human creativity has recently emerged as the soft underbelly through which a wider world of commerce and finance can be penetrated. The intentions of big media, big tech, and big money are provocative to the point of deliberate confrontation. They’ve already been caught stealing intellectual property in order to ‘train’ a.i., and they’re quite unapologetic about it. They are, after all, interested in control, not literary freedom.

The venality is utterly brazen. Why pay royalties to a writer of generic fiction if a machine-mind can cherry-pick the human imagination? The resulting bot-books may contain echoes of Atwood, but you can’t copyright an echo. That is why there is an over-arching sense of impunity. Offended creatives will have to spend years picking their way through a labyrinth of logarithms if they want to prove that the latest bodice-ripper from HAL 3000 is an amalgam of their stolen work.

This is not to say that writers, photographers, painters, actors, dancers, musicians, singers are powerless and disposable, but they are replaceable as vassals of big media. If you work in any of these roles you will already be worried, but you should probably worry harder and act now to ring-fence your creativity.

In my automated performance piece, I’ve chosen to focus on the human voice, one of the first targets in the war on personal creativity. We all own our own voices, but text to speech tools are becoming more sophisticated all the time. I went looking for ‘Jouliette’, my protagonist, and found her on a subscription platform alongside several other characters. Their voices were free to use, but the text needed considerable alteration to make them remotely expressive.

I also sourced copyright free images, music loops, and video from a variety of sources, and  an early-career musician put together some original music. About a half of the visual content was generated by myself using only my iphone. Similarly, the iMovie and Garageband software on my mac-mini did most of the heavy lifting. All of this cost me only my time because I already possessed the hardware I needed to live my life.

‘Why Do Humans Sing to the Stars’ is essentially a demo. I would much rather see it developed as a theatre project with actors playing machines, rather than machines in place of real people. It is also important for me to get the idea out there quickly, before the creative landscape becomes awash with hand-wringing narratives about artificial intelligence.

I happily admit that it’s a provocative piece, but I hope will stir the synapses. I certainly want to mix things up because I think the creative response to a.i. has been a bit fatalistic. A good actor can take the human voice and fill it with depth and emotion, but a.i. is fast accumulating each and every voice that ever uttered a spoken word. It already sounds authentic enough to the passing ear, especially in voiceovers, links, adverts and, dare I say it, fringe events. Worse still, it is appropriating performances and a myriad other intellectual properties in their entirety.

At the moment, the creative professions are concentrating on securing blanket royalties and residuals from big media. Quite how is another matter. Disney has already moved to convert real extras into virtual-actors with all the ruthlessness of the Cybermen. There is some acknowledgement that a.i could be useful in the more prosaic areas of the performing arts such as marketing and audience development, but it’s a Trojan Horse in the making. Once inside, a.i. will infect every part of theatrical culture and heritage.

It doesn’t matter if my piece has any artistic merit or not, the point is that it was done on a ‘no budget’ basis. If I can do that with limited resources then think what multinational media companies can do. Closer to home, the “trusted” BBC has been busily developing its own replacement career-killers, including an a.i. voice that is alarmingly natural and, importantly, has an authentic regional accent.

My feeling is that professional membership organisations such as PRS, the MU, the NUJ, and Equity ought to be demanding deep consultation from government on the use of a.i. by big media. There needs to be a measurable distance between recycled art and truly original creative processes. At the very least, there should be transparency wherever a.i has been used extensively in creative projects. Personally, I would start with the BBC, which makes news about its a.i. programme publicly available on one of its least conspicuous web pages.

I think this sort of action would force the issue for funding of the arts at every level because the aim must be to ring-fence human creativity from the insidious advance of machine learning. The latter is driven by big money, but the arts, impoverished by comparison, simply need levelling-up money.

Joined up campaigning won’t be expensive compared to private sector investment in a.i., and it will be worth every penny to protect the integrity of human expression. In this, I would include sound engineers, lighting technicians, copywriters, freelancers, and other public-facing roles. The whole conversation needs to move away from replacing redundant humans and focus on command and control of the machines, otherwise the machines will be the ones in charge.

Michael Stephen Clark