Prototypes – No.17
Your idea needs a test to find out where it's going. You probably do, too.
Hey! I’ve been gone awhile. I’ll be back on a regular cadence soon.
In the meantime, people keep encouraging me to share more work-in-progress. That’s this. Consider it a Cool Stuff Update that points to a lesson I would have never learned on my own.
If you want to make something you love, make a prototype first.
Because you’re going to see it’s limits – physically from building the tool, and functionally from using it.
Building shows you what you actually have to do next. All the while, you develop your taste, because you’re actually putting your idea to the test.
And that…that’s a whole lotta fun.
This note is about building out one of those ideas – a recording pencil – and what it reminded me about my life.
Growing up, one of my dad’s rich friends always had these weird new technology gadgets. Night vision goggles. Glasses that make you feel drunk when wearing them. The Nintendo Virtual Boy. My mom used to laugh anxiously at the price tags when he’d mention them. We weren’t really well off, so it was part WTF, part must be nice.
Meanwhile, I thought it was the coolest thing in the world. Every time I’d see one of his new purchases, all I was thinking about was what if I made something like this? (Spoiler: that hasn’t really changed.)
So recent times have been kind of incredible for me.
In the past month, an indie hacker named Avi Schiffman wowed an audience with his Tab. Rewind – a competitor of Heyday – twitter-announced their Pendant. Humane showed off their long-awaited AI pin on Naomi Campbell at a fashion show. Meta revealed a new pair of AI Ray-Ban's that can record from your eyes, and, most prominently, don't look ridiculous.
These are all recording devices that you have on-hand all the time. They’re connected to your knowledge (or social) graph and, because it’s 2023, there’s an AI that lends context to what you’re staring at.
I geek out over things like this. Beyond the idea of new tech, the blend of dedicated memory tools and AI is squarely in my wheelhouse. I’ve basically made this my career.
What's most interesting to me is that the form of these tools is not a phone. They’re something else – not designed to solve everything like your iPhone, but dedicated to their function – your memory.
I have always wanted one of these.
If I'm alone, I like to talk out loud. It helps me think. When I'm with people, I love to talk it out. That really helps me think.
My past efforts tried to solve this with phone apps and a notebook. But I often forget to open my phone to hit record. It’s not always-on or designed solely to record, so there’s friction.
A universal recorder can capture the spark, though. The initial idea. The buildup to the outrageous laugh. Sometimes you say it in the perfect way, and the source material is saved for what comes next.
So, yeah, these tool announcements are great.
And I had no expectation they were coming five months ago when I decided to build my own.
I thought I knew exactly what I wanted – an audio recorder built into my pencil – to allow for instant recording while I jot notes.
The form was important for me, too. No one wants to use a pencil with a brick tied to it. So ChatGPT helped me find technical specs and size limits for a mic on a chip to fit inside my pencil. I had a clear read, so I went to work.
Gosh was I wrong. In very obvious fashion, I screwed this up. And this is why we build prototypes. They show you what you assumed and embarrass you about it.
Okay, enough deliberating. Here’s the prototype.
Now this is a prototype in every way – hacked together from another device’s mic, wrapped in black cardstock to match the metal ferrule. At this point, our aim was function over form.
The pencil is my preferred form for a recorder because I take a notebook everywhere. A pocket notebook for walks. A midsize for sits. I'm always writing a note, reflecting, mapping out work, doodling…something.
This Blackwing pencil is short – half the size – which gave us extra length to work with for the circuit board. All Blackwings have removable erasers, so our constraints were clear. Fit the device into the metal base of the pencil without being too long.
And it works.
My recordings start with the click of a button. Right there in my hand. Off with another click. I plug the extender into my computer to download the recordings, send them through a Whisper transcription, and push them into Roam to find tags.
Replacing my phone with the pencil click feels far more natural, too. When I write, it's right in my hands. The friction is mostly gone.
And that was my hunch. I built this to test that.
Now there's a ton I dislike, too. The form factor is a touch flimsy. Storage is super limited and hard to access. I figured that would be the case.
The biggest surprise was in the building process. I found and bought the parts, but Donna reminded me that our first priority is testing the function. So we bought a spy pen with similar dimensions, extracted the circuit board, and jammed it into the pencil.
Turns out, the audio quality of this spy pen is shit. But now I can simply replace that part and build on top of it. I know what I need to do next. Trust the process.
A great Paul Graham quote points at this prototyping process in startups.
The reason to launch fast is not so much that it's critical to get your product to market early, but that you haven't really started working on it till you've launched. Launching teaches you what you should have been building. Till you know that you're wasting your time.
At Heyday, we’re a startup building a thing that doesn’t exist, so, inherently, this relates. Our CEO, Samiur, calls this getting to the shitty first draft. It’s hard to remember when you want to get it right, but it’s the only path. Share it early and learn. The world will tell you what’s wrong.
At home, I have Donna, the epitome of a prototyper. She is the reason I’ve learned this lesson.
I think nine different items in our house and garden are prototypes – a bench in my office, a keyboard extender, blinds, a trellis, an exercise machine for Pix, and on. Before she builds anything, she makes a shitty first draft.
This is illuminating and 100% necessary. Occasionally we're done – the blinds and trellis are good enough. Other times, it feels like we're back at square one with Pix. At this stage, Donna's aiming purely for function over form.
But her wedding dress? That was ultimately all form. It took deep care to create. The dress started as a prototype, but once she drew the lines, pinned the cuts, and put it on, she knew what had to change. She committed. She spent months standing over it, and crafted a piece she will never forget. Neither will I.
That’s how an idea evolves to find its best form.
It’s been close to a year since I dropped an update, but I am having the time of my life.
If I had to pinpoint why, I think it’s because I took a long time to prototype my time to learn what I wanted. My interests. Where I lived. My career.
Around 2016, I got the itch to do something big, whatever the hell that meant. In hindsight, I wanted to prove that I could make something I loved. I saw people I admired doing it. Why shouldn't I do it, too?
In the process, though, I realized I loved the work. Whatever it was – starting Career Mode, writing a Denzel movie script, working on startups – I loved trying. I'd do something small, test it, and learn where to take it…or not. Many times it died away, but I’d built up my testing muscle.
And then I found the tools for thought community. I read their histories in books and their presents in journals and tweet threads. I made some software tools of my own.
I realized I needed to dive a level deeper. I researched how people learn and synthesize together. I joined an AI memory startup. We’re building it. I’m still trying shit.
Through this, I realized the benefits of prototyping extend far beyond making things.
One of my best friends is trying to find his thing as we speak. He’s had a taste of interesting work and it’s changed his mind on the career he wants. Big bureaucracy sucks when you just want to make cool things.
But he’s discouraged. It’s not happening soon enough, and it’s maddening to him. I wouldn’t say he’s lost, but you know when someone you love is restless, looking for something else? That’s him right now.
Prototyping takes time. It’s not linear. You know the next step mostly because something went wrong the first time. But that doesn’t mean the future step will land you in your place. Only that this place isn’t it. That’s where I was years ago.
I don’t know at all how it’ll turn out, but I’m on my path.
My first step was building a conversational AI onto my Google Home. I called it Project Socrates. It recorded my rambles on design and asked me to provide my rationale. It was pretty terrible.
But today, it’s a template that runs in my Roam when I start writing a new spec. It asks me questions about my design patterns and how I’m putting something into practice. It’s saved my ass a few times.
Now I’m hacking together recording pencils at night and trying to augment your memory during the day. Who knows what will come of this.
The point is, you can’t know what you actually want until you try it.
It’s why Paul Graham says the real work begins after you launch your product.
It’s why the work I’m doing at Heyday now is so much fun.
It’s why Donna’s built a home of half-crafted functional prototypes.
Because once you build it, you find the real insight.
It’s way more fun to learn from something you build, because you ultimately find what you want to commit to.
Then you build something you love.
For that, I’m so grateful to Donna for showing me the way.
Special thanks to Donna Figenshu for reading and revising earlier versions of this piece.
You are the epitome of the growth mindset, Brendan. Love this piece. Such great inspiration to share early and learn!