I became interested in AI the day I discovered I no longer had to transcribe recorded interviews. This most laborious of writerly tasks could now be done in minutes. A few months later, the process took seconds – with virtually no amendments. Soon after that, the software wrote summaries of lengthy, complex prose – work that I had previously been employed to do.
Dots joined, data centres grew, pennies dropped. The transcriptions became instant. Applications broadened: my home bristles with smart sockets tracking electricity usage; friends post generative images; I ask ChatGPT inane questions.
Other than the obvious impact on climate breakdown – anti-data centre activism is growing in response to current expansion – and the workforce, the greatest threats are unknown. Andy Burnham is currently recruiting for an expert adviser to lead him through these changes. No sector will be unscathed, yet the creative industries are especially vulnerable.
When I mention my concerns to a tech colleague, who has successfully automated multiple processes for serious financial reward, he suggests training a bot to write in my style – a horrifying prospect. A newspaper I contribute to sends revised commission agreements for all journalists: we must not use AI. Etsy, former bastion of the handmade, sends a similar email. This one is vague and non-committal – perhaps they should have asked ChatGPT to proofread it.
In music, AI is being used in artistic, interesting ways. Pianist Zubin Kanga performs Philip Venables’ piece Answer Machine Tape, 1987 – exploring the life of New York artist David Wojnarowicz – using augmented instruments to enable the keyboard to simultaneously type text, while composer Emily Howard’s work
DEVIANCE incorporates neuroscience via AI.
But it’s also an enormous threat. AI relies upon ingesting existing work to imitate it – there is currently no clear framework for consent, compensation or transparency around the systems.
