there is so much to be done that has to be done by people, and so many people doing little or doing a lot of little real value.

large language models have not changed this basic situation. we miss this because our current economy, lethargic with incrementalism, has a lot of people doing automatable work less valuable than less automatable work they might be doing.

'Meredith,' some guys ask, 'why won't you shove AI into Signal?'

Because we love privacy, and we love you, and this shit is predictable and unacceptable. Use Signal ❤️

Behold: @redmonk's Top 20 Languages Over Time and the decline of the Stack Overflow tags that feed the analysis.

Our analysis dates back to 2012. We correlate cumulative language usage as seen through non-forked PRs on public GitHub repos against questions asked on Stack Overflow. It's an incomplete story, but gives us a pulse on language traction using publicly available data.

Our problem is that the veracity of Stack Overflow as source data is increasingly questionable. Tags for the top 20 programming languages in the last period were roughly 10% of the volume of questions at Stack Overflow's peak.

We don't have an answer yet, but for now we continue to see the value in both tracking the trends over time and tracking the state of the underlying data.