that PRD you got from ChatGPT is probably garbage

the "Package Deal" fallacy is wrecking your app builds

SK
Sean Kochel

the "Package Deal" fallacy

There's a popular "thing" or way of arguing that people fall victim to… it's called the "Package Deal" logical fallacy

It's where someone assumes that things grouped together must always be grouped together

You see it a lot in politics, actually:

"If you're a [insert political party] that means you must believe [commonly held opinion]"

YOU: But Sean… you talk about vibe coding, you're not about to get all political are you?

ME: I'm glad you asked! No, no I'm not :)

People commit this type of fallacy with their app builds all the time

They often get some sort of PRD out of the LLM, and then just roll with it without asking many questions

It usually looks something like this:

"Hey Claude, here's my app idea - build it into a PRD with features and blah blah blah"

Then you get some PRD out the other side, tell it "make sure its an MVP!", and assume that's good enough to start building from

But that is a very, very bad idea

In my PRO Framework, I talk a lot about the "P" stage (Planning) and how it's really the linchpin of the entire system

If you don't front-load important decisions you usually end up with what the internet lovingly refers to as "AI Slop"... and that's mostly because you're letting the LLM make the decisions, instead of… ya know - the human trying to actually build the thing

Anyways… the whole idea of having a narrow "MVP" can be a bit… vague

That's why I was stoked to come across this Github repo from the CEO of Y-Combinator:

https://github.com/garrytan/gstack

Because one of the features it has is helping you define something called "the wedge"

So… what's a wedge and why does it matter?

Back in World War 2, when the Allied Forces invaded the beaches of Normandy, that was kinda like a wedge

Once they pushed through the beach, they had expanded access to the whole of Europe, which they could start taking piece by piece, "win" by "win"

So Normandy was the wedge that gave them entry into the "market"

And you want to build your MVP narrowly around one wedge

The best shorthand for it is this:

What is the narrowest thing you could build that someone would be willing to pay you for, and they'd be sad if they couldn't have it anymore?

What's the tiny part of a big market where you can win?

I did an entire Youtube video on this "wedge skill" and a few others this week:

https://www.youtube.com/watch?v=MM320sAhFoY

Here's a few famous examples to illustrate the point:

Stripe - started as a 10 line Javascript script to help developers do payments more easily

Wedge? ✅

Doordash - started food delivery in underserved suburbs, not nationwide

Wedge? ✅

Reddit - link sharing site for tech/indie-hacker oriented people

Wedge? ✅

Moral of the story:

Focus all of your effort on the wedge, and then expand based on customer feedback


WHAT ELSE I'VE BEEN READING ABOUT OR DOING

Things I’ve Learned

1.

Agent memory feels like it's still a big gap in the "prosumer" world we all exist in.

There's two pieces to it (or maybe more):

First, I'm tired of these Agents pulling up random facts from a conversation I had 4 months ago that's no longer relevant and framing an entire question around it. We should have more control over agent memories

Second, AI coding tools have similarly poor long-term context that I would actually want it to have

If anyone has a good recommendation for these things, please reply and let me know

2.

Everyone's been yammering on about how "Apple and Google are banning vibe coded apps"

→ it's interesting to me how quickly untrue/false narratives spread like wildfire across the internet

Are they going to put up more safeguards against buggy code and poor security? Probably, yeah.

Are they limiting updates to vibe coding PLATFORM apps (i.e., something like Replit as a mobile app)? Yeah, they don't want a tool that can generate and execute arbitrary code. That's always been a policy

A helpful tip to live life by: anytime someone tells you something, think "hmm… maybe" & then investigate the claim instead of sitting with existential dread

3.

What do you want to want?

Most people are wildly unhappy deep inside because they have unbounded desires and wants that aren't actually meaningful

The desire for power, riches, & fame can be quite destructive to the human brain - this article outlines why & what to do about it:

I put it all in a word doc if you wanna read it

I had to go through the way-back-machine to get this article because the site has since been hax'd

That's it for this week, see ya'll next time!

-Sean

mail

Stay Ahead Of The Curve

Join 15k+ entrepreneurs & creators building valuable stuff with AI vibe coding, agents, & automation.

mail

No spam. Unsubscribe anytime.