I build an app, Agentsy, and some of my customers with ADHD tell me it helps. I also see a lot of AI apps marketed straight at the ADHD community promising things they plainly cannot deliver, and it bothers me. Every time one of those ads turns up in my feed, I picture the person who buys it, tries it, watches it do nothing for the hard parts, and concludes they're the broken one. They're not. The marketing is.

So this post covers the actual scope of what an AI voice notes app does, mine included, and what no app fixes. It's going to be short and blunt, and it will probably cost me some customers. Good. If you read this and decide Agentsy isn't for you, the post has done its job properly.

What Agentsy (and similar tools) can help with

I'll start with the genuine mechanisms, stated precisely, because vague benefits are how this whole space earned its reputation in the first place.

  1. Capture friction reduction. Speaking is faster than typing, and opening one app to talk beats opening another to type. If your specific bottleneck is losing thoughts between having them and recording them, this genuinely helps. It's the one thing the whole category is actually good at.
  2. Loop-closing automation. If the voice note turns into a drafted email sitting in your inbox, you don't have to remember to write it later. The draft is already there. That closes the loop drawing on far less executive function, which is the scarcest resource on a busy day.
  3. Fewer decision points. One capture destination means you're not deciding, mid-thought, which app to open or which list the thing belongs on. Fewer decisions, faster capture. The decision point is usually where capture dies, so removing it matters more than it sounds.
  4. Passive surfacing. If the app drafts an email or creates a task, it puts the follow-up back in front of you without you having to remember to go looking for it. You don't have to trust your memory. The deliverable comes to you instead.

These are real effects, and they help with a real symptom: capture friction. That's the honest size of the benefit. Now the honest size of the limitation.

What Agentsy cannot help with

Also specific, also honest, and this list is longer, which tells you something.

  1. Task initiation. If you can't start the thing you need to do, an app that helped you record the commitment doesn't help you do it. You still have to open the draft and hit send. The draft is easier to face than a blank page, and that's worth something, but it is not the same as starting.
  2. Time blindness. The app doesn't help you know how long the follow-up will take. You'll still underestimate it. So will I.
  3. Motivation. No amount of drafting a follow-up makes you want to send it if it's a hard conversation. The words can be perfect and still sitting there nine days later.
  4. Rejection sensitivity. If you're avoiding the email because you're worried about the response, automation doesn't touch that. The draft can make it marginally easier to face, and that's all it can do.
  5. Focus. The app helps between periods of focus. It doesn't lengthen the periods themselves. Nothing I've built touches that, and I'd be wary of anything claiming otherwise.
  6. Executive function in general. The app reduces the cost of one specific task. It doesn't increase your overall capacity. You still end the week with the same amount of you.

What no app can do

Here's the bigger picture, and I'd rather point at what the researchers and clinicians who've spent their careers on ADHD actually say than at my own product page. Russell Barkley's work frames ADHD as a disorder of self-regulation and executive functioning, which is exactly why an app that helps with one capture task can't treat it. Ned Hallowell has said for years that the foundations come first. The things with actual evidence behind them:

  • Medication, which helps the majority of people who try it, roughly 70% by most reviews, and which no app replaces or imitates.
  • Cognitive behavioural therapy adapted for ADHD, which works on the patterns apps can't touch.
  • ADHD coaching, if you can access it and afford it, because external accountability does what internal structure can't always manage.
  • Sleep, exercise, nutrition. The unglamorous foundations that matter enormously. No app touches these, and every honest clinician puts them near the top of the list.
  • Structure: external, visible, redundant. Calendars you actually look at, lists in places you'll actually see them, systems that don't rely on remembering.
  • Community. Talking to people who get it. This does more for the shame than any tool ever will.

Apps are at best a small assist alongside some of these. They are not a substitute for any of them. If you're struggling and haven't spoken to a professional, please do. That's not a marketing line. It's what will actually help most.

Why I'm writing this

Plainly, then. I could sell more Agentsy subscriptions by being vaguer about all this. I don't want to. The AI-for-ADHD space is going to get uglier before it gets better, and the community is going to, rightly, get more sceptical of it. I'd rather be one of the honest ones and sell fewer subscriptions to people who wouldn't have benefited anyway. An app that doesn't help you is worse than no app, because you end up blaming yourself for the failure, and I've read enough abandoned-system graveyards to know exactly how that story goes.

If you want to try it anyway

Small closing, no fanfare. If capture friction is your specific problem and you want to try Agentsy, it's free while you're testing. If it doesn't help within two weeks, delete it. That's the honest recommendation, and it applies to every app in this category, mine included.

The author does not have ADHD but wrote this after conversations with customers and readers who do.