Article

AI as a productivity app vs AI as an executive-function prosthetic

By Lachie · CyberFreedom · Updated 29 July 2026

The difference is what each one assumes about you. A productivity app assumes your executive function works and helps it do more: organise tasks, plan the day, write faster. An executive-function prosthetic assumes the opposite, that the part of your brain that starts tasks, weighs options and holds state runs unreliably, and stands in for it mechanically when it fails. If your real problem is "I have a system, I just can't get myself to use it," no productivity app has ever fixed that, because it's the wrong layer. The prosthetic works at the layer that's actually broken.

So when people ask "what's the best AI for ADHD," the honest answer is that the best app matters far less than which of these two jobs you're hiring AI to do. Pick the wrong framing and the best tool in the world does nothing.

What's the difference in one line?

A productivity app helps a working executive function do more. An executive-function prosthetic stands in for an unreliable one. Same AI, completely different job.

That's it. Everything below is the consequences of that one difference.

How do the two framings compare?

Here's the same AI, the same model even, used two ways.

The row that matters most is "requires you to be motivated." Every productivity system I tried over thirty years (Getting Things Done, time-blocking, Pomodoro, deep work) failed at the precise moment my motivation wasn't there. They were all built for a brain that produces activation energy on request. The prosthetic works because you don't have to want to do the thing. You just have to start the conversation, and the conversation takes the first action.

Why does "best AI for ADHD" miss the point?

Because the question is about the tool, and the useful part isn't the tool.

I've used Claude, ChatGPT, Gemini, DeepSeek, Grok and a dozen others, heavily. For the first year and a half I used them like everyone else: a smarter search engine, a code helper, a brainstorming partner. They helped, in the same way supplements helped my ADHD. Useful adjacent to the problem. Not the thing.

What clicked, late 2025, was that AI is not a productivity tool. It's an external executive-function module I can call into service when my internal one isn't producing reliably. Same way glasses aren't a vision tool, they're an external optical module I bolt on when my eyes won't focus. The reframe sounds small. Operationally it changes everything.

So the better question isn't "which AI." It's "which job." Once you know you're hiring it as a prosthetic, almost any capable model does it. What matters is knowing which prompt to run when, and trusting the prompt will do the job your brain was supposed to do on its own. None of it requires a new app, a new subscription, or a new system. The model you already pay for is enough.

What does each one actually do when I sit down to work?

Take the 9am freeze: you know the task, you want to do it, and you cannot begin.

A productivity app meets you one step past that gap. It'll help you organise the task into subtasks, schedule it, set a timer, track the streak. Every one of those is genuinely useful, and every one of them assumes you've already started. They optimise the part you can do and wave at you from the far side of the part you can't.

The prosthetic does the opposite. It doesn't help you organise the task. It writes the first three minutes of the work as if it were you, badly, in a voice you'll fix later, and hands it back. You paste it into the real document and edit it. You're now working, because you skipped the step your brain wouldn't do instead of being nagged across it. Editing is the easy mode of the same task, and your brain does it effortlessly.

That's the difference made concrete. One framing makes a working engine more efficient. The other one supplies the ignition when the engine won't turn over. Discipline assumes the engine starts. For a lot of us the engine doesn't turn over at all, and you can't discipline your way into ignition. A starter motor isn't more willpower, it's a different part.

Does the prosthetic framing ever turn back into a productivity trap?

Yes, and it's worth naming because it's the failure mode I know best. AI is catnip for systems-thinkers. An ADHD brain that uses AI badly can over-engineer faster than it ever could without one.

I've started "improving the prompt" and surfaced hours later with the most elaborate multi-AI decision architecture you've ever seen: five outputs that needed synthesising, so I'd build a synthesiser, then a different synthesiser, then a synthesiser to combine the synthesised outputs, hunting for the model that would finally hand me the perfect answer. At the end of it I had a magnificent machine and nothing written. The whole apparatus was procrastination in the costume of work.

That's the productivity framing creeping back in through the side door, optimising the system instead of doing the work. The prosthetic discipline is the opposite: the prompt is good enough the first time, take what it gives even if it's mediocre, and edit it in the real document. The output's job is to be edited, not to be good. If you find yourself perfecting the setup, you've stopped using the prosthetic and started building a productivity app for an audience of one.

So which should I use?

Both, for different things, and be honest about which job you're hiring for.

If your executive function genuinely works and you just want to move faster, the productivity framing is fine and the apps are good. If the thing that keeps happening is "I have the system, I have the plan, I just can't get myself to start it / choose / keep the thread," that's not a productivity problem and a better app won't touch it. That's an executive-function gap, and the prosthetic is the thing built for it.

Where to take this next

The quickest way to tell which one you need is to try a prosthetic prompt on a task you've been avoiding. The free Prompt Library on the site has the task-initiation one ready to copy, so you can run it tomorrow morning and feel the difference rather than take my word for it.

If the prosthetic framing is the one that fits, the full set is the ADHD AI Operating Manual: six prosthetics, one per failure mode, every prompt and every honest failure case included. Pay what you want from $9. I wrote it using the workflows it describes, which is the only reason a guy with a drawer full of unfinished projects actually shipped it.