Becoming a Superworker in a 12-person company

Josh Bersin, who studies how work and workforces are changing, coined a name for what AI does to a capable person: it makes them a Superworker. Not replaced, not automated away, but amplified, doing more, and more interesting, work than they could alone. The term usually shows up in writing aimed at large enterprises with thousands of staff and a workforce strategy to match. That is a shame, because the idea matters more, not less, in a firm of twelve people.

In a big company, one amplified worker is a rounding error. In a twelve-person firm, one person whose output has genuinely stepped up is a noticeable share of the whole business. The maths of small teams means the Superworker idea lands harder here than anywhere. Bersin's framing is deliberately optimistic about people: a Superworker uses AI "to dramatically enhance their productivity, performance, and creativity," which is a story about enlarging what a person can do, not about doing without them.

What a Superworker is

A Superworker is an ordinary skilled employee who has folded AI into how they work, so that the same person now produces more, covers more ground, and reaches a higher standard than they did before. The marketer who now also runs decent first-pass design and analysis. The operations lead who drafts in an afternoon what used to take a week. The account manager who walks into every meeting properly prepared because the prep assembles itself.

The key word in Bersin's framing is enhance. The Superworker is not a person doing the same job slightly faster. They are a person whose job has grown, because AI cleared the routine parts and freed them to do more of the work that needs a human. That distinction shapes everything about how you approach it.

Why it matters more in small firms

A small firm feels every constraint personally. There is always more to do than there are people to do it, and hiring is slow, expensive and risky. The usual answer is that everyone works harder, which has a ceiling and a cost. The Superworker offers a different answer: the same people, doing meaningfully more, because the tools have lifted what each person can produce.

This is also where a small firm's structure helps. There are no layers between deciding to work this way and doing it. One person can become a Superworker this month, and because the team is small, the effect is visible to everyone almost immediately. Greg Shove, who writes about AI and leadership, stresses that this only happens if someone actually demonstrates it, that leaders should model the behaviour rather than just endorse it. In a twelve-person firm, one person openly working this way is all the proof the rest need.

Picking the first person and task to amplify

You do not turn a whole team into Superworkers at once. You start with one person and one task. Choose the person who is already curious about the tools and good at their job, because amplification multiplies existing skill rather than substituting for it. A strong performer with judgement becomes much stronger. Pointing the same tools at someone who is struggling tends to produce fast, confident mistakes.

Then pick the task where amplification will show. Look for work this person does often that has a lot of routine wrapped around a core of real skill: research before the thinking, drafting before the editing, preparation before the meeting. Aim AI at the routine wrapper, so the person spends more of their time on the skilled core. That is where output visibly steps up and where the rest of the team sees what is possible.

Amplify a strong performer and they get stronger. Point the same tools at a weak process and you get faster mistakes.

Enlarging the role rather than cutting it

This is the choice that decides whether the Superworker idea helps your firm or quietly damages it. When AI clears half of someone's routine work, you have two options. You can take the saved time as a headcount cut, or you can let the person do more and better work with it. In a small firm, the second is almost always the right call, and not for sentimental reasons.

A small firm wins on relationships, responsiveness and quality, all of which are human. If you use AI to shrink the role to its old output with fewer people, you save a little and lose the thing you compete on. If you use it to enlarge the role, the same person now does the deeper client work, the better thinking, the things you never had time for, you get a real gain in what the firm can offer. Bersin's whole argument is that the prize is enhancement. A small firm that treats it as headcount reduction is collecting the smaller prize and giving up the larger one.

The honest limits

Amplification without judgement just makes mistakes faster. AI multiplies whatever it is pointed at, including bad habits and poor decisions, so a Superworker has to be someone who can tell good output from bad and is willing to check. The role of the human shifts toward direction and quality control, and that judgement is exactly what you must not let atrophy. Speed on top of poor judgement is not a Superworker. It is a liability with better tooling.

There is also a pace limit. Becoming a Superworker is a real change in how someone works, and it takes a few weeks of awkwardness before it clicks. Expect a dip before the lift, and do not judge the experiment in its first fortnight. The people who become genuinely amplified are usually the ones who pushed through the early clumsiness, not the ones who expected it to feel effortless on day one.

One more caution, specific to small teams. If the rest of the firm reads the Superworker as a threat rather than an example, the experiment curdles. The way to avoid that is to be explicit that the point is enlargement, not replacement, and to make the amplified person a teacher of what they learned rather than a quiet competitor. Handled well, the first Superworker becomes the route by which everyone else gets better. Handled badly, they become a reason for everyone else to keep their heads down.

What to do this week

Pick one person: curious, capable, already decent with the tools. Pick one task they do often that wraps real skill in routine work. Give them the time and the licence to fold AI into that task properly, and ask them to work in the open so the rest of the team can see it. Then, when the routine clears, resist the urge to bank it as a saving. Let them do more. One visible Superworker is the most persuasive case for AI a small firm can make, and it tends to make the next one volunteer.

Working out who to amplify first, and how to enlarge the role rather than hollow it out, is the kind of judgement the AI Lessons for Leaders sessions are built to develop. If you want one person in your firm working at a visibly higher level, book a personal lesson.

Sources and further reading

• Josh Bersin, "The Rise of the Superworker: Delivering On The Promise Of AI", 14 January 2025. Where the "Superworker" comes from: a person whose output is amplified, not replaced, by AI.
• Greg Shove, gregshove.com, 2025. His point that leaders should model AI use themselves rather than just endorse it.