Aiming at a moving target
Thoughts on AI and the labour market
It’s summertime in Northern Ireland. Which means BBQs that are often called off by rain and flies that can enter your house with the exacting precision of an incoming inter-continental ballistic missile, and as soon as they are in your house they develop amnesia and bob at the windows for several hours. Inevitably, they’ll start to pester you, and you’ll reach for something to swat them with. At which stage you’ll discover that hitting a moving target is actually quite difficult.
This week I was in a conversation with a bunch of business leaders, public servants and policy makers from across the UK. We were discussing the government’s AI adoption priorities for business (and tangentially, the government’s own AI adoption priorities).
The conversation started with the question: how can AI be used to augment what people are doing rather than disrupt it? It’s an easy question to nod along with approvingly, but I’m worried that this question is swinging the swatter at the fly that’s already on the move.
The conversation took a polite trajectory, with no one (myself included) really grasping the nettle of the potential awkward conversation. I’ve revisited the question several times since then and have decided that the premise of the question might be unhelpful. It’s a question really masquerading as an answer. Perhaps a better question is what does work become as we enter a new industrial revolution?
I think what disturbed me most is that augmenting work assumes that the nature of work isn’t going to change. In fact, I think, the concept of work is going to be an increasingly moving target. So, in that case, if work is going to keep moving then our interference needs to be in trying to direct where we want it to go.
I always come back to the tweet from Joanna Maciejewska. The question for me is how can we shape the coming disruption to allow humanity the freedom to focus on the things that are lifegiving.
When we minimise the argument by trying to be overly precious about protecting what is, we don’t make enough room for creative destruction and positive disruption.
It’s not that AI can’t augment humans in work (it should, in fact!), it’s more that trying to avoid disruption feels like a surrender to potential progress. So many jobs have reached a natural conclusion. I haven’t seen any Indeed postings looking for knocker-uppers, gas lamplighters, lift operators, scriveners, bowling alley pinsetters, or video store clerks.
But jobs ending is only half of the story. Because jobs don’t always end. Just as often they mutate, metamorphose, and re-form. It’s interesting that everyone assumed that the ATM would be the end of the bank teller. Instead, banking teller numbers climbed for the best part of 3 decades after the introduction of ATMs. It turns out, banks became less expensive to run, so banks opened more of them as they expanded into new territories and offerings. The job rewrote itself around the things that the technology couldn’t do. Offer advice. Support a confused customer. Build a relationship. The banking clerk didn’t disappear; instead, the work was fundamentally rewritten.
I’m not trying to paint an overly rosy picture, either. There is a proliferation of AI slop, and there are the questions about scaling technology with this potential, both environmental and ethical. It’s going to be disruptive, but we have a chance to shape the disruption.
Shaping, however, is an active word. It means choosing, together, which parts of work are worth protecting and which we’re happy to let go. It means facing into some tough choices for how work is going to be disrupted in the age to come. That’s a hard conversation. Much harder to answer, for example, than “how did that damn fly get in here?”