Quick reminder…

3 days ago we opened Bestapps Academy: our private community where we help you build and launch real AI software and get your first user.

10 founding spots are already gone. 40 left.

$29/month, locked in forever, plus exclusive perks and access. After the 50th member, the price goes to $69/month.

Join Bestapps Academy as a founding member →

Ok, let’s get into it.

Most people upgraded everything this year: new models, new agents, new automations stacked on top of each other.

Then they sit down and type instructions to all of it at 40 words per minute.

You speak at 150 (almost 4x faster).

Every prompt you type is running at a quarter of your actual speed. You upgraded the whole stack except the input.

So here’s the solution:

You don't send a colleague a three-word brief. You explain the context, the constraints, what you've already tried. But typing all that into ChatGPT takes forever — so you don't.

Wispr Flow lets you speak your prompts instead. Talk through your thinking naturally and get clean, paste-ready text. No filler words. No cleanup. Just detailed prompts that actually get you useful answers on the first try.

Millions of users worldwide. Works system-wide on Mac, Windows, and iPhone.

Try Wispr Flow free

Quick history check. The keyboard layout you're using was designed in the 1870s. For typewriters.

When personal computers reached the masses in the early 80s, they had one problem: machines couldn't understand a word you said. Speech recognition barely existed. So computers borrowed the typewriter's input, and everyone moved on.

Typing didn't win because it was better. It won because voice wasn't an option.

That constraint is dead. Today's models understand your rambling, half-formed spoken thoughts better than they understand your carefully typed bullet points. More context, more nuance, more of what you actually mean.

But the habit survived the constraint. Most people still funnel every thought through their fingers, 40 words per minute, like it's 1986.

Typing is the last manual step left in your AI workflow, and easily one of the easiest ones to upgrade.

1. Brief by talking

Stop crafting prompts. Talk them.

Open Wispr Flow (or your phone's dictation) and explain what you want the way you'd explain it to a colleague. Rambling is fine. Repeating yourself is fine. The model structures the mess. And a 300-word spoken brief beats a 40-word typed one every single time.

2. The rambling draft

Got an idea for a post, a pitch, an email? Don't stare at the blank page.

Talk for 3 minutes. Dump everything, in whatever order it comes out. Then paste the transcript into Claude and ask it to find the structure. You get a draft in your own words in a fraction of the time.

3. Capture on the move

Your best ideas don't show up at your desk. They show up on walks, in the car, between meetings.

Voice memo → transcript → your notes. 30 seconds instead of a lost idea.

Dictating feels awkward at first because you try to speak in finished sentences. Don't. Five habits fix it:

Stop performing. You're not recording a podcast. Ums, restarts, "wait, actually", the model handles all of it. Speak like you think.

Say the constraints out loud. When you type, you cut the context to save keystrokes. When you speak, there's no cost. Say the audience, the format, the things to avoid. That's the 80% your typed prompts were leaving out.

Give it a destination. End every ramble with what you want back: "turn this into an email", "give me 5 versions", "tell me what's missing". One take is enough. The model sorts out the rest.

Make it one key away. Set up a hotkey for dictation (Wispr Flow has one, so does your OS). If starting takes more than a second, you'll fall back to typing without noticing.

Know when to keep typing. Exact names, short commands, editing a specific line: keyboard. Anything longer than two sentences: voice. That's the whole rule.

Here's what a spoken promptt actually sounds like:

That's a terrible piece of writing. And a near-perfect prompt. Audience, tone, constraints, the tricky nuance, the ask, all of it in 20 seconds. Try typing that context out and you'll quit halfway.

Stanford ran the numbers back in 2016 (yes, the study is literally called "Speech Is 3x Faster than Typing"). English input: 3x faster, 20.4% fewer errors. And that was ten years ago, with speech recognition way worse than what's in your pocket right now.

Now do the math on your own day. If you spend 30 minutes typing prompts, messages, and notes, speaking cuts that to 10. That's 20 minutes back, every day. Over 120 hours a year.

Three full working weeks just from changing your input.

Your next 3 tasks: speak them. Don't type a single one.

It'll feel weird for about 5 minutes. Then you'll notice your prompts got longer and richer, and the outputs got noticeably better. Worst case, you feel silly talking to your laptop for a day and go back to typing. Best case, you never go back.

Stefano Martell
Co-Founder, The AI Leverage by Bestapps.ai

ps: Learn how to master AI and build useful, monetizable stuff here: https://www.skool.com/bestapps-academy-8162/about