Striking while the AI is Hot

13 Aug 2023

Given the impact that ChatGPT has had on the world since the end of last year, I'm not sure that I would choose right now to resort to strike action if I were a writer of TV drama. ChatGPT is an AI-powered language model that can be trained to write human-like text based on context and past conversations - in other words, by mimicking anything that's ever been written on the Internet.

ChatGPT has proven to be surprisingly good at writing like a native speaker. This has caused consternation to many, including tutors attempting to grade their students' coursework. Some institutions already ban the use of AI to produce work for submission, but if certain students are dishonest, how can the invigilators tell?

The quality of what ChatGPT can generate is undoubtedly impressive. However, for all its cleverness, it does have a rather curious flaw which even the creators of its software can't fully explain: its ability to hallucinate - a technical term used by the AI world to refer to "making stuff up". Now this is not a trait you want to see in an AI entrusted to generate news items, reference material, or any other factual prose; but if you were to give it the task of writing dramas for someone like Sky, I think that might be a different matter. In fact hallucinating might be actively encouraged,

On May 2 this year, the Writers Guild of America (WGA) which represents 11,500 screenwriters, went on strike. One of the main focus points in the labour dispute is the residuals from streaming media. Given that media streaming is seriously big business right now, an unscrupulous producer might decide that AI is less hassle. And if they did, how easy would it be to prove it?

I think if I were were a writer of TV drama right now, I'd be trying to prove that you just can't get better than the real thing.

Go back