The Corporate Playbook for Upskilling a Truly AI-Bilingual Workforce

These days, the term “AI-bilingual” keeps appearing on boardroom decks. It sounds a little awkward, much like “digital native” did fifteen years ago, but the idea is to do actual work within businesses.

The promise is simple: a worker who comprehends the nature of their work as well as the rhythms of collaborating with machines that increasingly perform certain tasks. The way it is executed is completely different.

You can see what I mean if you stroll through the learning and development area of a big bank in Canary Wharf. Prompt examples are scribbled on whiteboards. A few mid-career managers squinted at outputs while hunched over laptops, arguing whether the question was the issue or if the model misinterpreted it. People in fitted suits are learning to type commands to a chatbot like first-year graduate students, and there’s a slight, almost embarrassed energy in the room. The generational mix is difficult to ignore. Younger employees can be impatient and move quickly. Older people are better at asking questions.

Most businesses started with the incorrect premise. They believed that this was a technology rollout, which involved purchasing licenses, conducting webinars, and cleaning up after themselves. It hasn’t been successful. The companies that are actually making progress approach it more like teaching a second language, with all the repetition and patience that goes along with it. Internal certifications have been conducted covertly by Goldman Sachs.

The focus of Unilever’s redesigned learning portal is “augmented work.” Even a Cleveland regional hospital network that I learned about last spring trained nurses in a brief course on documentation tools, and the administrators who funded it were taken aback by the outcomes, which included improved triage notes and fewer night shift errors.

Senior leaders are the target audience for the six-month Cambridge program, which I frequently hear mentioned at conferences. In the end, it promises not only literacy but also an implementation strategy. Really, that’s the trick. It’s more important to understand where a model will subtly break or fix things in your business than it is to understand how it operates. Investors appear to think that businesses that succeed in this will gain a significant advantage over the next three years. Perhaps they are correct. It’s still unclear if the productivity gains will end up where the consultants said they would or if they will disappear into quiet headcount reductions and reorganizations.

The executives I’ve spoken to seem to believe that technology isn’t the greater risk. The middle manager is the one who subtly guides their team around it because they don’t fully understand or trust it. A few well-known companies have already suffered from uncontrolled, careless AI use, including leaked prompts, client data hallucinations, and a humiliating customer service exchange that went viral. Training programs are still plagued by those tales. Recently, I’ve noticed that every curriculum has a lengthy, somewhat defensive section on risk and ethics.

Curiosity is something that none of these programs can truly impart. This is what distinguishes employees who pass the certification but forget it by Christmas from those who become truly AI-bilingual. This round is likely to be won by the businesses that figure out how to reward the curious as well as the technical. The others will continue to purchase licenses, host webinars, and ponder why the transformation never materializes.

Being employed in any white-collar job at this time is peculiar. One awkward training session at a time, you get the impression that the playbook is still being written as you watch this happen.