In the early days of computing there was a vision that one day, everyone would build their own software. AI-coding tools have arguably removed the most alien limitations of this for most people.
But the truth is, writing code was only one part of the problem.
Most people don’t make their own clothes, invent their own recipes, investigate their own breaking news. Not because the tools don’t exist to do so, but because of a truth discovered through human history:
When we specialise, everyone benefits.
The most underestimated aspect of products is that we are buying the taste of the creator, as much as their ability to make the thing.
And so, the hope that this technology gives me is that instead of a digital world shaped by the homogenous technical culture of Zuckerbergs and Gateses – it will be equally shaped by journalists, philosophers, women and minorities.
It’s time for heterogeneous software.
example
Let’s take an example:
I want a tool that lets me transcribe YouTube videos and podcasts and pull out key insights.
It’s very very easy to get generic tools that do this.
I don’t want a generic tool. I want an opinionated and specific tool. I don’t want it to include stupid baseline observations. I want it to skip those, or link elsewhere to where they are covered already and tell me something I don’t know.
I want it to identify any books or other products mentioned in the transcript and give me links to them in their own section.
I want it to crawl and find other interviews that dig further into the aspects I’m interested in.
I want it to be really really good and free up time I waste listening to audio in real time.
I think it’s the combination of easy dissatisfaction and opinionated desire for problem-solving that leads to the creation of products.
This should be a boom time for curation. I want tools built by people I trust, and it’s easier than ever for those people to build multiple tools influenced by their opinion.
But that’s very different from everyone making their own software. And I think that’s ok.