Andrew Palmer, Group Editor
Meeting A Theakston Old Peculier Shortlisted Crime Author: Jo Callaghan
What are the benefits of AI? Does it interfere with the crime process, or do you think it genuinely helps?
As quick as in the blink of an eye, the title of her novel on the Theakston Crime Novel of the year short list, she replies, “AIDE (Artificially Intelligent Detective Entity) Lock can do some things very, very fast in a more evidence-based way, but in other areas there are real weaknesses in terms of if you have experienced a crime.
"I've experienced a break-in; such incidents are distinctly human and need another individual to bear witness to your suffering and distress." While AI is obviously incapable of doing many tasks, it can technically perform some tasks, such as facial recognition. However, most AI facial recognition systems are trained on data from largely white populations, ethnicities, which can lead to racial bias."
Instead of viewing AI as merely cheaper and quicker, Callaghan wanted to start a debate about how it can enable us to have better lives and a better criminal justice system.
"I would like AI to do my bins,” she quips.
"There are potential strengths and opportunities, but also weaknesses, and the issue is to work out where humans add value and where AI could enable us to work better and smarter, rather than just having an unthinking adoption of it. Just because it is technically possible doesn’t mean we should just do it.”
That's why Callaghan created her crime series: "I was carrying out research into the role of AI in the health workforce, and the impact on different professional roles, which led me to the possibility of creating an AI detective."
"I started researching and discovered that the police force was already utilising algorithms and other AI tools in both police and judicial processes. Research has raised data-related concerns, such as the potential for AI to perpetuate racism and misogyny due to the data it is trained upon thereby reinforcing existing prejudices. I was so excited about the multitude of ideas that emerged. That’s what galvanised me to write this novel.”
Callaghan is quick to point out that she is trying to take a neutral stance with her writing and refers to lots of potential benefits. Arguably, using AI features could help make decisions more transparent and explicit.
Although she had the idea for the book, Callaghan couldn’t write it because her husband was very ill. “I wrote it two months after he had died,” she tells me, to keep herself sane.
Jo Callaghan
"I wanted to say something about humanity to show how we all struggle with life to capture love, humour, and hurt."
Interestingly, Callaghan tells me that in terms of crime novels, she likes authors who write with warmth, compassion, and humour, like Kate Atkinson or the late Susie Steiner.
“I try to make my books humorous in part as well as thought-provoking.”
So, I ask how much planning she does.
“I spend a lot of time thinking about my books; my finger-to-keyboard ratio is quite low. When I'm cleaning, walking, or just doing other things, I am cogitating about the plots, so when I do sit down at the weekend, I can be quite productive.”
Callaghan considers herself lucky. When she started writing, she thought it would really be a Marmite book because it is set a small number of years in the future, and it is about AI.
“When it came out in January 2021, it was just a few weeks after ChatGPT took off; I was lucky with the timing as it went from being a very speculative book to quite a now book.”
The BBC chose her book for Between the Covers because it resonated with many people's interests in and concerns about AI.
"Hopefully, it won't be frightening," Callaghan says. "It's not too technical; it makes it interesting because the story brings out all the issues that we were talking about in a narrative form."
"We've all got to engage in these debates so we can shape the future and the way that AI develops to enable us all to have better lives."
“We are all capable of learning from each other. There are things that AI can do really well, but equally, there are things that only humans can do. It's not either/or it's trying to figure out where each adds value and what we can learn.
“The key question at the heart of the story is if AIDE Lock is capable of deep learning, how much is he capable of, and what happens when he does?
“With all the great shows, like The Terminator or Spock in Star Trek, you never quite know to what extent they are developing consciousness or human qualities. So you might infer. Spock may act in a way that initially leads you to believe he is beginning to learn, but then he may act in a way that leads you to conclude otherwise.
"What I'm trying to toy with is the extent to which we impose our own human senses and assumptions about development upon them, while also understanding that to be human is to be at the apex of the development tree. It's not always clear how much AIDE Lock is actually learning, as opposed to simply learning how to say the right thing.
“None of us ever know what the other person is thinking, whether they are AI or human.”
If that's not challenging enough, Callaghan informs me that at the beginning of the book she use a quotation from F Skinner's Contingencies of Reinforcement: A Theoretical Analysis:
“The real question is not whether machines think but whether men do. The mystery that surrounds a thinking machine already surrounds a thinking man.”
.jpg)