Product

Following my previous cast about new product experiments. We've been building something new for the past week. It's called ⌘ INTERN. 

1/ What is this? ⌘ INTERN is an experiment in AI-powered content surfacing.

2/ Problem: Most founders are too busy building to document their journey. Not because they lack insights, but because they lack a system to capture them.

3/ Solution: AI that runs alongside you, turning daily work into meaningful content—no extra effort required.

4/ Progress: 5 days in
• Tested 5 AI models
• Generated 59 high-quality pieces
• Running this very experiment through the system

5/ Follow along as we build
Everything you see tagged "Surfaced by ⌘ INTERN" is content the system found in our logs.

About a month ago we had an idea to solve our own problem — getting better at sharing progress and building publicly.

Last Friday we got the first MVP working. It looks bad, but feels so freaking good.

Little thread about what's been happening these last few weeks

the art guy™ co-founder @lisafound, ex r/ga. notes on art and design. screenshots of things i've learned. all misspells are my own.

Naturally I started sharing my method with friends, sending them prompts, getting on calls to explain how to set it up and asking for feedback. This is where I got frustrated. I was updating prompts more than I'd planned and it was hard to share. Things kept breaking for friends and I didn't know how to fix them. And most of all I wanted to get their feedback and it was hard to collect.

So we decided to build a small MVP. Use Vercel API, connect Anthropic, do simple no-code website and go. We've been wrong. Nothing worked. Things broke all the time. Some days we felt like we're almost there, others felt like going back to square one.

Then it just worked. Our dev made his usual spells, threw some magic and last Friday MVP was up and running. We played with it for an hour, made more posts than ever before. Now we have MVP on our hands.

We spent 0 time thinking about design or brand, and this is quite intentional. Strong brand creates high expectations and we can't afford it yet. Things will break and that's okay. Our goal right now is to find people who would want to use even this crappy version because they need it, not because it looks good.

Then it just worked. Our dev made his usual spells, threw some magic and last Friday MVP was up and running. We played with it for an hour, made more posts than ever before. Now we have MVP on our hands.

We spent 0 time thinking about design or brand, and this is quite intentional. Strong brand creates high expectations and we can't afford it yet. Things will break and that's okay. Our goal right now is to find people who would want to use even this crappy version because they need it, not because it looks good.

https://cmdintern.framer.website/

Our website is free Framer template, we didn't even pay $5 to turn off Framer logo. Our design system is out of the box. Little design thinking that’s been done was prompted in v0 in about 10 minutes. We just need feedback. I know lots of things are bad, but better way to say is I assume they are bad. I'd like to hear this from people who need this product.

One thing we did is started Twitter channel (Farcaster soon), so ⌘ INTERN can post its own content about its own creation. Feels like dogfooding our own content to the system is a good thing to show how it works. Again many things could be improved, but that's later.

Our website is free Framer template, we didn't even pay $5 to turn off Framer logo. Our design system is out of the box. Little design thinking that’s been done was prompted in v0 in about 10 minutes. We just need feedback. I know lots of things are bad, but better way to say is I assume they are bad. I'd like to hear this from people who need this product.

One thing we did is started Twitter channel (Farcaster soon), so ⌘ INTERN can post its own content about its own creation. Feels like dogfooding our own content to the system is a good thing to show how it works. Again many things could be improved, but that's later.

https://x.com/cmdintern/status/1904610848800198976

Right now we're only interested to find 10 people who struggle with content and want to give this prod a chance. So we can get feedback. We don't even have payment system yet to charge them, but it's coming soon.

All and all I'm beyond happy. It looks bad but feels so freaking good. Personally I feel like it's 100x from how we've done it last time, and ultimately that's the only thing that matters. Thank you for reading this and I'm likely will cold DM you later.

https://open.substack.com/pub/ilyyya/p/the-quiet-ai-taking-notes-on-your?r=if28l&utm_medium=ios

Our problem isn't novel or new. We like building stuff, but we forget to share. We wait for the perfect moment to announce, but a week/month/year later we have no idea what we wanted to share. Since I'm the one working on social media, I was using lots of AI to help myself. It works, but I've noticed that I have to set up my context for AI every time, collect data, etc. It's annoying and I fail to do it consistently.

I decided to spend some time and solve this with a no-code prototype. Turned out really well. My goal was to teach AI to analyze my docs, find interesting things to talk about and suggest content in my voice in various formats. I worked on prompts, collected data, set up a project in Anthropic so I can jump in whenever and don't need to worry about context windows.

Co founder @Intract - private data stack for AI & crypto use cases. Documenting my journey to become 100x AI first founder. Growth nerd, part time guitar player

i love this. i was discussing this exact idea yesterday & wanted to build it out <3