Transmission ·4 min
Stop saying no, just build it
I went to Costco twice yesterday, three times total in the last two days. I help my friend run big supply runs and deliveries for cargo ships here in Honolulu and we fill up three SUVs in the process, exhausting ourselves but having a lot of fun in the process.
It takes a long dang time to fill up one of these orders! There are three of us shopping at Costco at any given time, each one armed with a flatbed cart and a customized list that my friend puts together by hand, printed out and color-coded for each of us.
That’s a lot of work, and I think I just might be able to help out here because I have Fable and I have unlocked my brain…
Unlocking your brain is the theme of the Fall Cohort and, if you’re curious what that means, you can read more here.
I could make a simple task list… or…
This is where the notion of “leverage” really takes hold: when you’re thinking about what’s possible with AI programming. I know I can crank out something fast, but I also know that, with a little more effort and tokens, I can crank out something magnificent.
I think I did just that! Behold, my latest app: Flatbed:
This my “Costco Speedrun” app, which divides the Costco floor into one, two, or three vertical “stripes”, or zones. Each person has a zone and a list, divided by department, and a progress bar that shows them where they’re at compared to others.
Because this isn’t just a “let’s get through Costco fast” app, it’s a friggin RACE with a lot a stake! If you win, you get a WAFFLE PARTY!
If you lose, however…
I’m still working on the finger toys…
It happened overnight
My friend started this whole thing off by asking Claude on her phone to color code a list of items based on a spreadsheet. The color codes corresponded to “zones” within Costco where each of us would do our shopping. Unfortunately, when she tried to share her list, it didn’t work due to privacy and security.
That’s when I thought: “Opus 5.5 was released today… let’s take this thing for a spin and see what i can do.” Next thing you know, I’m adding a waffle party for the winner and virtualizing my friend’s 20 year-old Gen Z daughter into an AI bot that heckles us…
This couldn’t have happened without a frontier model like Opus 5.5 or Fable 5.1. I gave it some reasonably specific directions, but then added “amaze me” with every prompt. I also told it to use /high-end-visual-design from the Taste skill (go get this if you don’t have it) and, full honesty here, I had it done within 3 hours.
Yes, you can
The amount of fun you can have with AI in programming is bonkers, if you change your thinking just a wee bit. For me, the thing I have to keep telling myself is to “stop saying NO dammit!” because things are truly that easy these days.
The other day I was thinking I needed a CRM tool and started looking at Pipedrive, Attio, Notion, and Trello (among others). They were either extreme overkill or far, far too simplistic. Then it hit me that it was probably a lot easier just to make my own.
45 minutes later I had something that was far superior to anything I had seen that day, and it could grow with me in the exact way I needed it to (a CRM is a well-known thing, so building one with Fable is very simple). This is where we are now: you can build it, you just have to do it.
It has to be outstanding, however
The problem that every SaaS has right now is that Fable can recreate what they do in a week’s time. That’s neat and all, and it will save you money if your goal is to replace something you’re paying too much money for with something you create in a weekend - but that won’t save your company or your job.
You have to be outstanding and build applications that are sensational pieces of art! You have the design skills - just go online and pick one! You have the coding practices so you don’t write garbage too; now comes the fun part: building something that no one has built before.
Or, better yet, the thing you’ve always wanted to have, but just haven’t had the time. But that’s only the beginning, friendo, because the thing you’ve always wanted is rooted in the ideas of today…
Where we’re going, we don’t need any roads!
The key to AI: leverage
AI is great for speed, and it’s wonderful for efficiency, planning, and execution. That’s not the real strength, however. That comes from you.
By combining your inspiration and creative mind together with the brilliant technical capabilities of a model like Astra or Fable, you can create something that will absolutely freak you out. This is what we’re going to focus on in the cohort, so if you find yourself at a loss for what that looks like, you’re not alone.
Seats are filling up, and if you want in to a creative AI jam session where we build what we always thought was impossible, click this thing:
I know so many people are wondering what’s next, and I see that as an opportunity. We get to define what’s next for ourselves!
In my next note, I’ll talk about what this process looks like in more detail. Until then, thanks for reading, and yes I typed this by myself, in my new list server that I once thought was impossible to build. It even has a slop meter attached!
I’ll tell you more about this app in the next note too.
