I’ve just been to an ALCS (Authors’ Licensing and Collecting Society) focus group meeting. Amongst everything else, we debated AI and what we authors ought to be doing about it. Unless you’ve been living under a rock or travelling in the past, you have to be aware that this element of the future has been rushing towards us recently. There was a large debate about it at Eastercon too, and I was on that panel. As I write this, for example, I’m watching another of those YouTube videos that has AI generated images – extraordinary things.
There’s an author planning to ‘write’ and publish 10,000 novels this year. That’s the accurate number, I’ve not hit a few zeroes by mistake. 10,000, which, if he works 9-5 with an hour off for lunch, 5 days a week, works out at one every ten minutes. It cannot be possible to even read the novels he’s ‘writing’.
This’ll swamp the market with cliché ridden stories. Is that what we want?
(They are cliché ridden as they take all of literature that’s been stolen for them and average it for their output.)
Of course, they will improve until eventually they will be ‘better’ than humans can write.
I’d suggest that, as a species, we should be deciding what we want to be doing. Is it that machines produce art and writing, and we all do the tedious jobs for greedy rich people?
Art with a big ‘A’ is, to my mind, human-to-human communication. Remove one side of that and it’s not. Part of being human is expressing ourselves. I have stories I need to tell. AI threatens to take that away from us making us less human. (Mind you, we’d need AI to read 10,000 stories a year, so maybe AI will be writing for AI, and we’ll all be sitting on the patio drinking Pimms.)
I suggested at the focus group that ALCS ought to campaign that writers should be compensated for the use of their writing as material poured into these AIs’ dataset.
“You need to run the campaign and save civilisation and the human race,” I said. Slightly over the top, but it is all the stuff of science-fiction. For my part, the back of my latest novel (Dawn and Dave of the Dead – https://amzn.to/3JOJ50h) has a jokey “☑ Not written by a robot” on the back. Perhaps it is needed.
P.S. This blog was also not written by a robot.
Neither is this response written by a robot.
I fully endorse David Wake’s fears about AI and its impact on all aspects of art. Take away the stories written by people, then – as David asserts – it’s not art.
The other fear that I hold concerns AI and its role in software development. During my time as a university lecturer, I helped to advance the skills of undergraduate software engineering students, sending them out into the world equipped to develop software for the benefit of the public and private sectors. Perhaps the Post Office/Fujitsu scandal is an isolated instance of blind faith in the correctness of compute output? I don’t know; I doubt it. I wonder how long we will have to wait until an AI system produces output that is difficult to challenge because “the computer is always correct”. Blind faith is AI systems could lead to cases similar to the Post Office scandal.
It took several years to show that the Horizon software produced incorrect output, during which time many sub-postmasters and mistresses were wrongly prosecuted: the consequences were hugely damaging to these innocent people.
Before I close this response, I note that the Post Office has announced an extension of Fujitsu’s contract for its Horizon software, despite the lengthy probe into the scandal. Blind faith still persists in this case. This announcement beggars belief!
David Muir, 24th July 2023.