Benchmaxxing aside, if you are using those tools for programming on a regular basis it should be self-evident that they are improving. I find it very hard to believe that someone using LLMs today vs what was available one year ago (Claude Code released Feb 2025) would have any difficulty answering this question.

Even if your goal is to go out and create value for others, your contribution is proportional to what everyone else can offer. If others with AI will deliver that value cheaper, or if what I am good at can be easily automated, it's getting harder and harder to deliver more value than I consume.

If anyone can go in, take a GPL project like chardet and reimplement it using LLMs, then the current maintainer just saved everyone time by making their implementation publicly available.

Our legal framework wasn't built for a situation where reimplementing complex software is trivial, much less almost completely automated.

OpenAI and Anthropic have ~40 billion yearly revenue already, and growing fast. I don't know if 10x that is the right amount to invest, but I'm pretty sure 0.1x of that is not going to cut it.

European AI investment plans are always so timid. This article mentions a couple of investments on the order of $1.5-2.5 billion USD that their government says "would position the UK as a world leader in AI". Hard to imagine you can be a world leader in anything when investing two orders of magnitude less than your competitors.