Querying pypi to fact-check AI hype - by Jonathan Bennion
Azure/AWS growth (peaked again at the beginning of Q4 2024) normalizing to 97% / flat respectively indicates enterprise AI buildout cooling
Google's steady 31% baseline growth suggests market shift from hypergrowth to stability, but this is also loosing ground quickly in Jan 2025.
AI hype is still rampant - I’d queried github activity in August 2024, and this was before US government spend kicked in at the beginning of their fiscal year in Oct 2024, which gave us another [late] peak in spend / AI usage, further propagating hype-fueled AI gypsies globally. US government spend was misleading for growth, but also predictably behind the curve due to a large government moving slower than average, so how can that be parsed from any resurging enterprise growth (if exists, and as hyped)?
More realistically, what data can tell us what is real? Another way to find actual package usage is through PyPI package downloads - actual insight to adoption patterns of major AI platforms so you can compare it to any hype that states otherwise.
