Which decisions should we delegate to AI is a much discussed question at the moment, and an important one. Far more pertinent than the ‘will they wipe us out’ question.

It’ll be no surprise that I’ve been thinking about it for a while, and hopefully no surprise that I think the answer is nuanced and a bit personal.

When I was preparing for the Creative Futures project with the Defence, Science and Technology Lab (part of the UK Ministry of Defence), I mapped out all the interconnected aspects that might affect our futures which were swimming around inside my head. I came to think that there are three overarching considerations for society, and one of those is around the extent to which we delegate our agency and autonomy to technology.

This was also something that I touched on during my keynote to the Robotics and Autonomous Systems Showcase in the summer this year. Using the story, Zygosity Saves the Day, from the project with TwinsUK as the example. One of the themes in that story is the delegation of health prevention via a personalised health plan created by an algorithm and an automated care home for the poor, not the rich.

More recently, I’ve been writing an article for the BSFA Focus magazine titled, Embracing the Chaotic, where I explore what we should and shouldn’t delegate to the machines. Without giving any spoilers, I talk about my need to keep some chaos in life and to not be totally ‘efficient’ in everything – I need the space to be surprised. But, of course, not by some weird random health cure!

So, I reckon the degree to which you delegate your decisions to the algorithms, the AI, depends on your circumstance and context. Critically, we need to know when decisions are being made on our behalf, on what basis and by whom or what. That way we keep our agency and make best use of the emerging technology at the same time.