Guides · How-to

How to turn voice memos into a to-do list (without copy-pasting into ChatGPT)

Voice memos are the easiest capture tool ever made. You're driving, walking the dog, or lying awake at 2am, and the thought arrives: book the MOT, reply to the electrician, Maisie needs her PE kit Thursday. You hit record, say it all, and feel briefly lighter.

Then the memo joins forty-seven others called "New Recording", and none of those tasks ever reaches a list. The capture worked. The processing— listening back, splitting it into tasks, putting each one somewhere it will actually get done — is the part nobody has time for. That's the step this guide fixes. Here are three ways to do it, from most manual to least.

Method 1: The manual way (Voice Memos + Reminders)

If you record in Apple Voice Memos on a recent iPhone, you already have transcripts. This works, costs nothing, and needs no new apps:

  1. Open the recording in Voice Memos and tap the transcript icon (the quotation marks) to see the text.
  2. Tap the three-dot menu, then Copy transcript.
  3. Read through the wall of text and pick out the actual tasks. A two-minute ramble usually hides four or five of them among the thinking-out-loud.
  4. Create each one by hand in Reminders (or whatever list app you use), adding the list, date and person yourself.

The problem isn't any single step — it's that the whole loop takes longer than the recording did, so it only happens when you have a quiet ten minutes. Which is exactly what you didn't have when you reached for voice in the first place. Most people do this diligently for a week, then the memos start piling up again.

Method 2: The ChatGPT way

The popular workaround is to put an AI in the middle: copy the transcript, paste it into ChatGPT (or Claude), and ask it to "extract the tasks from this as a bullet list". It's genuinely good at this — it will find the five tasks in your ramble and ignore the part where you were narrating traffic.

You still have to do the shuttling, though:

  1. Record the memo.
  2. Copy the transcript out of Voice Memos.
  3. Open the chatbot, paste, and prompt it.
  4. Copy the bullet list back out.
  5. Paste it into your to-do app — and unless your app imports lists, that still means re-typing them as individual tasks.

Five apps-and-pastes per memo. It also loses everything around the task: the AI doesn't know your family, your business, or that "the electrician" means the quote you've been chasing since March. You get a generic list that you then have to categorise and assign yourself. As a one-off rescue for a long memo, it's great. As a system you run every day, it leaks.

Method 3: The one-step way (a voice-to-task app)

The third option is to collapse the pipeline: record into an app whose whole job is turning speech into tasks. The capture and the processing become the same step. If you're evaluating apps in this category, look for four things:

  • One recording, many tasks.Real brain dumps mix topics. The app should split "dog food, the MOT, and can someone ring Mum" into three separate tasks, not save one blob.
  • Sorting that uses your context. Categories should come from your actual life — your family, your business, your words — not a fixed menu of Work/Personal/Errands you have to maintain.
  • Capture from anywhere. Siri or Action-button support matters more than it sounds — the thoughts arrive when your hands are full, and unlocking-finding-opening is where capture habits die.
  • Somewhere for other people's tasks to go. This is the one most apps miss, and it deserves its own section.

This is how Offload works: you tap record (or use Siri or the iPhone Action button), say everything in one go, and each recording arrives as separate tasks, already categorised using your own context. No transcript to copy, no prompt to write, no list to re-type.

The step nobody talks about: half those tasks aren't yours

Listen back to any real brain dump and notice how many items are actually for someone else. Ask Sam to book the dentist. The PE kit — that's a Thursday-morning job for whoever does the school run. The invoice needs to go to the accountant.

Every method above, done in a personal to-do app, ends the same way: those tasks land on yourlist, and you become the reminder service — forwarding, texting, chasing, checking. The list got organised; the load didn't move.

That's the gap Offload is built around. From the inbox of sorted tasks, you send each one to its person — it lands in their Offload app with a push notification, on theirlist, not a shared one. You see Sent, Seen and Done for every task, so "did you get my text?" stops being part of your evening. There's more on why transferring ownership matters in our guide to delegating the mental load.

Which method should you use?

  • Voice Memos + Reminders — fine if you record a memo or two a week and everything on it is yours to do.
  • Transcript + ChatGPT — good as an occasional rescue for a long, messy recording; too much friction as a daily system.
  • A voice-to-task app— the right answer if voice is your main capture habit, and the only answer if the tasks need to end up on other people's lists too — a spouse's, a teammate's, or an assistant's.

Whichever you pick, the principle is the same: a voice memo isn't a to-do list — it's raw material. The system that wins is the one where processing it costs you nothing.