The Translation Tax
You’re not burned out. You’re misrouted.

You take what you actually think — the precise observation, the pattern you caught three layers down, the concern that is almost certainly correct — and you convert it into a format the room will accept.

You soften the edge. You add the preamble. You read the energy before you speak. You decide which version of your thinking will land without triggering the wrong response. You perform fluency you don’t actually feel.

Then you do the meeting. Then you process the meeting. Then you debrief yourself on whether what happened was what it looked like, or whether you misread something, or whether the thing that seemed fine was actually not fine.

By the time you get to the actual work — the judgment call, the analysis, the decision that needed your full cognitive capacity — you’ve already spent a significant portion of it. Not on the work. On the overhead around the work.

That is the translation tax.

The distinction that matters

I want to be precise about something, because this argument is easy to misread.

Emotional intelligence is a real leadership requirement. Reading a room, calibrating communication, understanding what someone actually needs from an interaction — these are not optional skills. They matter. I am not arguing against them.

What I am describing is a difference in how that work gets done.

For most neurotypical leaders, the social processing layer runs automatically. It happens in the background — below the level of deliberate thought. They walk into a room and the read happens without a conscious decision to read it. They calibrate tone in real time without allocating attention to the calibration. The emotional intelligence functions like a built-in AI agent: always on, low overhead, largely invisible to the person running it.

Which means it draws from the same cognitive pool as everything else I’m doing at the same time.

This is what I hear consistently from other autistic leaders. Not universally — autism is not a monolith — but often enough to recognize it as a pattern rather than a personal quirk. The social processing is not absent. It is not even necessarily slower. It is effortful in a way that, for many neurotypical peers, it simply is not.

The tax is not about lacking emotional intelligence. The tax is the energy cost of running manually what others run on autopilot.

And that cost — sustained across a full workday, across years of leadership — is where the depletion comes from. Not from incapacity. From the processing overhead of doing consciously what the environment assumes is automatic.

The misdiagnosis

Most people are misdiagnosing what the tax is doing to them.

The distinction matters because the fix is different. Rest does not solve a routing error. You can step back for two weeks and return to the same system and be depleted within days, because the system that misrouted you is still there, unchanged.

What the tax looks like

You have a meeting at 2pm. Before it, you spend fifteen minutes drafting a message to clarify what kind of input is expected, because the agenda doesn’t tell you and walking in uncertain has a cost. You spend another ten minutes thinking through how to frame the concern you need to raise so it doesn’t get dismissed before it’s heard.

You’re in the meeting. You’re doing two things simultaneously: following the content and monitoring the room. Both are running on the same cognitive resources.

After the meeting, someone asks how you think it went and you’re not sure, because you were processing so many signals at once that the overall read is still compiling.

You go write the summary. You spend time on tone.

That was a one-hour meeting. The translation overhead — before, during, after — was another ninety minutes of prefrontal cortex work. And none of it appears on a productivity dashboard.

The routing solution

The reason this matters now, specifically, is that AI has made the routing decision visible in a way it wasn’t before.

When I built the AI Cognitive Strategy Matrix, I wasn’t mapping how to use AI tools more efficiently. I was mapping a decision architecture: for every type of cognitive work, where should it go?

Translation work — formatting, phrasing, social overhead, summary production — sits in the low-stakes execution quadrant. That is Q4. Cognitive offload territory. Not because it’s unimportant, but because it does not require your unique judgment. It requires execution. And AI handles execution.

What AI cannot do is the Q2 work. The scenario mapping. The signal interpretation. The judgment call where your specific way of seeing — the pattern recognition that runs differently in you than it does in anyone else in that room — is exactly what produces the insight no one else caught.

For neurodivergent leaders, the translation tax is highest for a structural reason. The environment was not designed for your cognitive architecture. It was designed around a default. Which means the interface cost — the overhead of operating in a system not built for how you think — is higher for you than for the people around you.

This is not a deficit. It is a design mismatch. And design mismatches have design solutions.

When you offload the translation work — not just some of it, systemically, as a routing rule — what comes back is not just time. What comes back is the cognitive capacity that should have been yours all along.

The question I kept asking myself was: what would I think with if I got it back?

The answer, once I built the routing system, was more than I expected.