Hi, I'm Peter. This is my website.
Tech's meritocracy problem goes political.
Working with coding agents can create intense dopamine rushes. But what remains after the initial high wears off decides if agentic engineering is just the next high-tech slot machine or something genuinely useful.
Can't we just be happy for one moment without overthinking it?
Writing your own software is more than just customising - It's your first step into taking control of your computer-life back
How to get back control over your own data, one custom software tool at a time.
Looks like it's now a very specific kind of software that is doing the eating.
Agents can erode the current producer/consumer hierarchy in software engineering and create new ways of participation and software creation.
Agentic engineering can dramatically shift the current model of software customisation, and with it, the current model of participation in software in general.
Letting the system do all the creative stuff is fun and all, but the real power comes from systems that enable you and your growth.
When I started to automate my own processes, I quickly noted that there are two kinds of automation: One that removes the tedium of everyday tasks, and one that removes the ability to learn and grow. But which is which?
Introducing AI in an organisation is not just about improving processes or results. It's also about transforming the organisation itself.
Many organisations treat AI adoption projects as "innovation lab" work. And with that framing, they miss the real opportunity.
How we see and interpret the actions of others can tell us a lot – Mostly about ourselves.
Balance feels like something static. This is misleading.
If most of your codebase is written by AI agents, will you still understand your own product? And: Does it matter?
