- A speaker on Y Combinator's podcast argued that building is now easy and knowing what to build is the harder part. It was an opinion backed by anecdotes.
- One speaker described their own move from plain files to a database. It is one person's account, not a general finding.
On Y Combinator's podcast, a speaker argued that when software is cheap to make, the hard part is knowing what is worth making. Paul Copplestone, who runs the database company Supabase, was the guest.
What was said
A speaker described a group of "builders": people who are not software engineers but need something made. The speaker said they have pent-up ideas they could never persuade an engineer to build. Domain experts, who know the gaps in their niche, can now dig in without recruiting anyone.
One speaker described a fintech founder who had exited his company. He once needed an engineer for everything. Now he builds his own finance tracking, including bank links. The speaker said such software would not have been built even two years ago.
Copplestone told a second story. After Supabase's conference, a man from Brazil told him he had built a taxi app from his phone, a small Uber for his local market. Copplestone said the man is a taxi driver but he did not know why he built it.
A speaker drew the lesson: "the much harder thing to do is to know what to build you know being able to build is something that anyone can sort of have on tap now".
Early on, a speaker said builders using AI coding tools find that files only get you so far. Later, a speaker described their own personal app. They first wanted everything in plain Markdown files, then found a database and a screen useful. That is one person's account, not a general finding.
Why it matters
Our reading: the fintech founder fits the thesis best, because his banking knowledge shaped what he built. The taxi driver shows what is possible. Whether his app reflects deep market knowledge is not known.
A pile of files is hard to search across, hard to update safely and hard to restrict by person. A database is built for those jobs. The speakers did not list these reasons, so treat them as our gloss.
For managers, the questions follow. Who in your company is already building tools with AI? Where does the data in those tools live, and who can see it? Training may need to favour judging what deserves building over making it.
The other side
Copplestone runs a company that sells to these builders, so he has a commercial reason to be upbeat. The evidence is anecdotes. Nobody gave figures on how many such projects work, last or stay secure.
A speaker also said: "the number of people who know how to actually have an impact on society with it is like more or less constant". That was an impression, not a measurement. It sits awkwardly beside the talk of many new builders. The conversation left open whether judgment will spread or stay rare.
Written by the WebPulse Newsroom with AI assistance, and checked by our editorial review: every quotation was verified against the recording's transcript. How we use AI.
The conversation this talking point comes from
- Y Combinator: Why Everyone Is About To Have Their Own Database (2026-10-30)
