AI is compressing the time it takes to do good work. The problem is that most organisations measure effort, not outcome. When your three-week project takes four days, the question you get is not "great work" - it is "what are you doing with the rest of your time?"
Three months ago, a team lead at Intelegencia walked me through a project they had just finished. Four days. Same scope, same quality - the same type of project had taken us nearly three weeks two years earlier.
I said good work. Then two weeks later, someone senior asked if we had extra capacity on that team.
That question has stayed with me.
The career risk from AI is not that it replaces you. The risk showing up right now - the one most people are not prepared for - is that it makes your effort invisible at the exact moment your work has never been better.
Why AI Is Creating a Visibility Problem Most People Have Not Seen Coming
When AI speeds up your work, the people above you see the timeline. They rarely see what went into it.
The call that saved the project. The client conversation that happened before the work even started. The version you scrapped at 11pm because it was technically fine but would have landed wrong. None of that shows up in a four-day delivery. What shows up is four days.
TechCrunch reported in June 2026 that AI was the top reason cited for corporate layoffs for the third month running, with nearly 40,000 tech sector cuts in May alone. Marc Andreessen noted that most large companies are overstaffed by 25-75% and AI has become the excuse to act on it.
The executives making those calls are not looking at individual performance. They are looking at headcount against output. And that picture looks off when your team finishes in four days what used to take three weeks.
At Zendesk, I watched this happen before AI tools were this capable. A colleague who had moved into a more senior role got quietly marked as low output in a headcount review. His actual work was some of the sharpest thinking on the floor. What had changed was that he had stopped looking busy. Fewer emails, fewer meetings, fewer documents circulating. Just cleaner, faster decisions. He left six months later - not because he was pushed out, but because he could see it coming.
Speed is not the story. Don't let it become yours.
Why "My Work Speaks for Itself" Is the Worst Bet Right Now
It never really did. That was always a bit of a myth.
Harvard Business Review reported in June 2026 that middle managers are getting squeezed between executive ambition and day-to-day reality, with almost no support for adapting. The people most at risk are the ones carrying the most - because the strongest managers are absorbing AI-driven change quietly, without making noise about it.
Quiet competence has always been undervalued. AI is just making it faster.
A team lead on one of my accounts put it directly after I walked her through this: "So the faster we work, the more we have to explain why?"
Yes. Exactly that. It is genuinely unfair. It is also just how it works right now.
What Actually Keeps You Visible When Output Is Accelerating
The work that stays visible is not what you did. It is the decisions you made.
Why this direction and not that one. What risk you spotted that the data did not surface. What you told a client in week one that changed how the whole thing landed. These are the things no AI generates, and they also disappear from the record if you do not put them there yourself.
This is not about talking yourself up. It is about making sure your thinking exists even when you are not in the room.
At Intelegencia, I started asking teams to add a short note alongside each piece of work - two or three lines on what the key call was and why. Five minutes. It changes what a senior person sees completely. They stop seeing four days and start seeing what went into those four days.
I am still not sure this is enough. The way most companies measure contribution was built for a slower world. AI is moving faster than the systems designed to recognise it, and I do not have a clean answer for that gap. What I do know is that waiting for the work to speak for itself is not working anymore - if it ever really did.
