Vibe coding platforms are blowing up. But the data shows it's more of a creation vs. a consumption story.
Ex. @lovable_dev - traffic to Lovable itself is 3x greater than all traffic to sites published on Lovable 👇
People are making things for themselves...not the world (yet!)
These products will only get better as the models improve. And, there is clearly builder excitement!
I see enormous opportunity in product execution that will be key to "winning" the space - and hope to see vibe coding platforms hosting apps that themselves get massive traffic.
A few Qs I think are key:
- What constraints do you impose, and how do you communicate them?
- How do you tradeoff mass market appeal from expressiveness?
- How do you make it FUN?
- How do you build in "vibe deploying" and "vibe debugging" in a similarly delightful way?
• • •
Missing some Tweet in this thread? You can try to
force a refresh
1/ We’ve been living in a software shortage for 40 years.
Only a small fraction of people could code, so society underproduced the most elastic good in the modern economy.
Now AI is collapsing the cost of software creation.
This article was posted four months after chatGPT launched and it’s criminally underrated. It opens the door to a bunch of ideas that I think are important and rarely talked about. (more below)
2/ Every major technology wave begins when a scarce input becomes cheap enough to waste.
CPUs became cheap and personal computers emerged, bandwidth collapsed and streaming took off, storage costs plummeted and the cloud was born.
Today the cost of software production itself is collapsing which will release a wave of innovation.
3/ The essay asks why software has taken so long to “eat the world,” and uses a simple price/quantity graph to explain it.
When the cost of software is high we underproduce it, and when costs fall demand expands.
For decades we were stuck at the wrong equilibrium because of Baumol’s cost disease (hardware got exponentially cheaper while coding stayed human-labor-intensive) so expensive engineers meant too little software and mounting technical debt.
LLMs collapse those costs, shifting the equilibrium and unleashing new demand.
🧵💀 As @pmarca says “There is no substitute for knowing what you’re doing.”
I’m convinced that most product managers / designers have no idea what to build and have instead created a bunch of process to mask this failure. I count myself in this group at times in my career 👇🏽
So there are just two types of projects. Projects where a person who Knows What To Build tells you and everyone builds it. And projects where you have to hit a metric and it doesn’t matter what you build as long as you hit it. Let’s call them Type 1 and Type 2 projects.
The problem with Type 1 projects is that no one knows what to build. Most companies have zero people who have this skill but even if they do the person gets drowned out by the org.
Yep, we just invested in @ArcBoats the "Tesla for Boats". Why?
- Repeat founders + team from SpaceX, TSLA & Boeing
- Electric is a 10x for every part of boating: no maintenance, cheaper, quicker, and unlocks access to protected waters
-Aerospace design practices & A+ software
👇🏽
When I learned that two of the best engineers from Credit Karma (Mitch Lee) and SpaceX (Ryan Cook) were teaming up to build the "Telsa of boats", I was intrigued to say the least. Today, say hello to Arc Boats: arcboats.com
I knew Mitch (@dontmitch) from my Karma days. He could not only code, but also lead teams and inspire an organization. Proof point: he was the only non-executive named as a "key man" when Credit Karma was acquired by INTU.
1/n Financial services are often though of as a zero sum game - there's a fixed pie and therefore “winning” is finding clever ways to take a bigger piece of it.
But the intersection of tech and finance has actually been characterized by positive sum outcomes - a bigger pie!
2/n Let's talk about negative sum (where both parties are worse off as a result) v. zero sum (where one party must lose so that another can win) v. positive sum (both parties are better off) situations in financial services, and why the fintech approach is so important.
3/n Negative sum: A real world example is tariffs in trade - typically both countries typically suffer as a result of tariffs and barriers to trade.
In financial services: Debt settlement - the lenders and the borrowers both suffer.