If you leave them to it, the answer’s always “no”.
The need to have AI entrepreneurs embedded in teams to shake things up was just one of the learnings from the latest BusinessLDN and Deloitte AI SteerCo, which met last week. We took stock of what’s changed since we last met, conducted a deep dive into the upskilling and reskilling agenda, and tried to understand the workforce strategies needed to tee up successful AI adoption.
So, what’s changed since last time out?
The pace of change has accelerated sharply but unevenly, both between firms and amongst workers within firms. In the space of a few months, we’ve gone from AI predominantly being about taking out tasks to now taking out jobs.
The labour market has got more complex. On the one hand, the talent market has heated up, with access to the best talent getting harder as many of the big AI players move to London. Lots of poaching is going on between the big tech firms. Offers are being met with counter-offers. Some firms are observing that the graduates coming out now are better prepared than their existing workforce. And yet hiring is slowing at the entry level, with many firms choosing not to bring people in rather than letting people go.
The international talent pool is complex, too. It’s where some of the best AI can be found, but visa requirements can be costly. Some firms are thinking about how to make the best use of the global market and employing people where they sit, wherever that may be around the world. And yet often, clients want people “in market”.
And firms are becoming more conscious of weighing up the risk-reward balance. They are continuing to embrace the opportunities but are more alive to the risks, largely driven by Mythos. This means a stronger role for governance teams. There’s also more of an eye to the cost of tokens and ensuring value for money in the deployment of AI, so finance is also more to the fore.
What did we learn about upskilling and re-skilling?
Not surprisingly, one-size-fits-all, top-down approaches are doomed to fail. Similarly, we heard that self-serve catalogues are close to a waste of time – blighted by too much choice and rapidly becoming out of date.
So, what does work?
Peer-led sessions are very popular. Informal sharing of tips and good practice, organised locally and often run as lunchtime seminars.
Small beats big every time. Micro-modules that can be done on the move, with a maximum video length of 5 minutes. Think Duolingo meets AI training.
Keeping at it. Firms are spending lots of time coaching on prompts. Some firms are aiming for about 15% of working hours given over to learning, with the added rigour of quarterly check-ins on skills across teams. And make sure to give AI another go: things have progressed apace, so just because it didn’t work previously, don’t give up on it.
Finally, what were the early pointers on a workforce strategy to tee up successful AI adoption?
Workforce planning at a time of rapid change is hard to do, but the first tip is to make sure you have a workforce strategy. Try and do at least two to three-year planning horizons and think through what your skills requirements will be across the organisation. Then replicate this on a more detailed team basis.
Dive into the cultural reasons behind any lack of take-up. Look out for the “inefficiently busy”.
Firms are increasingly looking to appoint a Chief AI Officer to help with deployment.
It can be a good idea to involve a third party, both for their ideas but also to tap into their knowledge of what is working well in the marketplace.
Don’t just have a central team. If you leave it to governance and IT, the easier answer will nearly always be “no”. Have local AI entrepreneurs embedded within teams to agitate for change. And get them to enter into a dialogue to try and identify an architecture that risk and IT can get comfortable with.
That said, the governance question is critical. Data compliance is typically the key thing holding back use cases, and often for very good reason.
Despite the intensity of discussion about displacement, almost no one is yet doing large-scale lay-offs or designing support programmes for re-deployment outside of their organisations. All were clear that business has a big role to play in skilling up at scale, but government will have a role in addressing displacement.
The next meeting of our SteerCo will be focused on what it all means for future skill requirements and the changes required to train tomorrow’s workforce.