The workplace baseline is moving, and firms will have to teach their way through it – Schultheiss

That baseline is now moving to artificial intelligence (AI) tools, and the shift is better understood as a labour market story than a technology one, because the new baseline cannot be recruited. The pool of people who genuinely work well with these tools is small, familiarity is routinely mistaken for competence, and firms that want the capability at scale will have to build it themselves. 

We reached that conclusion at Uinsure and have spent two years testing what building it involves, from our leadership and senior teams through to several completed cohorts of graduates and young people, and the experience has produced lessons I suspect apply well beyond our own walls. 

Adoption is led, not mandated 

The first is that usage follows visible leadership behaviour, not policy. People change how they work when they watch the people running the business change first. 

I tested this on myself first, appointing an AI coach at the beginning of last year and setting a goal of completing no meaningful task without using these tools. It was harder than I expected, because established habits are efficient, but the effect has been real. The same practice now runs across our executive committee and has changed how we prepare for decisions and how quickly we reach them. 

Sponsored

Your First Home: confidence to build

Sponsored by Halifax Intermediaries

Only once the change had taken hold at the top did we give it a permanent home, establishing an AI Lab as a dedicated environment for developing how these tools improve the way we work. 

The sequence mattered because the Lab was the institutional version of a change we had already made personally, and it is where our structured training was born, first among our development teams, then the graduate programme. 

The tools are ordinary, the difference is depth 

There is nothing exotic about the tools themselves. Claude is the main one across our business and our engineers work with others that are readily available. When the technology is a commodity, the return on the spend is decided almost entirely by how thoroughly people use what they have, and depth is a matter of habit and confidence rather than procurement. 

The culture in which people reach for the tools instinctively, and understand where the boundaries sit, has to be built deliberately, and it is the only part of the investment a competitor cannot replicate by placing the same order. 

Structure changes the speed 

The lesson that has genuinely surprised us is speed. We expected structured learning to outpace individual experimentation. What we had not anticipated was how quickly people would progress from complete novices to confidently executing extremely complex tasks. As a training exercise, cohort members with no coding background have built test websites and platforms in areas relating to their personal areas of interest. 

An excellent young colleague, James Price, arrived through a traditional apprenticeship and now leads the training. Every session pairs that growing capability with safe use, including how the tools should be used correctly in a regulated business. 

The sessions also taught us about the anxiety beneath adoption. The most common question at the start of a cohort is whether the technology will take people’s jobs. 

By the end, the same people are talking about how much faster and further they can go with an AI companion alongside them. We did not argue anyone out of the fear. Hands-on experience resolved it, which suggests the answer to workplace apprehension about AI is more likely found in a classroom than in an internal email. 

What happens next matters 

Our biggest test started in September when the training started to be delivered to every colleague, because a baseline is only a baseline if everyone reaches it. If the entry-level skillset is changing, someone has to teach it, both to the people already inside our business and to the generation arriving into it. 

For a business like ours, which exists entirely to serve its partners, the case is even more direct, because the people developing this capability are the same people building and pricing what those partners deliver under their own brands. 

The technology will keep improving without any help from us. The people using it will not, and that is the part a business can actually do something about.