Seemingly, thrusting professionals are seeing AI as a fast track to partner, reckoning that all those tedious years of working your way up the hard way will now be taken care of by generative AI tools like ChatGPT. Hurray! We’ll all get rich quicker!
But privately these firms are rather worried. It takes ages for the academic curriculum to sync up with modernity, so what will happen to these shiny graduates with their googleable degrees, deposed as trainees by AI, but still expected to re-emerge as experienced hires just below partner level?
In school they’re already talking about the learning loss that occurs when kids outsource their knowledge acquisition to AI. This phenomenon isn’t new: we forgot how to remember the Odyssey when we learned how to write it down; we forgot how to hear music in all its fullness when we invented notation and tethered the scale to the equal temperament of a keyboard; and the introduction of calculators drastically reduced our ability to do mental arithmetic. But this might be the first time we have encountered a technology that can have that effect across so many categories of learning simultaneously. And if kids have experienced learning loss at school, then continue with their reliance on AI into university, will they come out knowing anything useful at all? If all they will have learned is how to use an AI brilliantly, well, AIs are already learning to be better at that too… Industry is rediscovering apprenticeships to make up for an already disappointing graduate cadre: but what do they need to start doing now, if even these apprentices will be usurped by AI?
Way back in 2003 at Ashridge, we initiated a research programme based on asking existing board-level leaders ‘what do you know now about yourself as a leader that you wish you’d known 10 years ago?’ The findings were used to devise a leadership accelerator and written up as the book Leadersmithing. Our research programme included collaborating with a neuroscientist to show how this kind of learning is acquired. From that, we showed the role of the emotions in learning, and found that reliable templates are most efficiently acquired through learning under pressure. And both this method and these research findings suggest an answer to the conundrum of workplace learning loss.
First, we need to get forensic about what, precisely, partners do, and how they learned it. It’s highly likely that much of their value-add is not AI-able, so this exercise should immediately reveal a workplace curriculum for those hoping to succeed them. The Leadersmithing list of critical incidents suggests it will be a fairly standard set of challenges, which will differ between workplaces and cultures only by degree and nuance rather than by type. For example, all partners will have had their mettle tested by making key decisions, fronting multi-million dollar pitches, and mopping up after things have gone wrong. And we know these things are teachable, if you can be precise enough about the muscle memory you’re trying to acquire, like practising difficult conversations or handling hostile media.
Second, we need to learn from the neuroscience. I remember answering the phone in my first ever London office, to hear my sibling, hiding in a cupboard at another London office, asking in a stage whisper: when you’re photocopying, do you take the staples out?! We all remember those ghastly days of learning the ropes largely by making mistakes and incurring the wrath of our seniors over everything from making the coffee wrong to sending out blank faxes. Life would indeed be tranquil if we could make AI take this pain for us. Our recall of such events is heightened by the fact that our errors were often observed. And indeed we learned vicariously, wincing at witnessing the mistakes of others, which is another argument in favour of a back-to-the-office policy. This is because our Ashridge findings showed that whenever you feel observed and under pressure, your heart-rate increases and your learning is enhanced, as the memories you form in those moments are stored deeply in your amygdala.
And we all learned far more than just office-craft in those clumsy days. Through the tedium of note-taking and bag-carrying we saw how leaders really behave: we learned about power, decision-making, values and standards. We witnessed the quite brilliant rescuing of an impossible situation, or a tension diffused with a beautifully timed witticism. We also learned how not to do it, too often I imagine. And it is this implicit learning that we now need to surface and teach back, so that we do not lose a whole generation to AI. Let’s use the gift of AI to remove the ritual humiliations of traineeship, but winnow out of it all the Leadersmithing we can find.