Your math is missing the fact that Apple products are differentiated from their competitors. If AI becomes a ubiquitous commodity, it's not worth 300B/y.
The point is that the market is big, and capturing a small part of it would make these companies wildly successful; and there's no need to make unemployment inevitable - I think this diminishes the AI Doom story quite a bit.
It absolutely diminishes the AI Doom story, but it also diminishes the AI Will Be The Golden Goose story, they are two sides of the same coin. The AI companies can only capture that surplus if there are not competitive substitutes available.
In the same way that Amazon gets rich taking ElasticSearch for free and charging for hosting, Amazon will take free models and charge to host them. The companies building frontier models have massive R&D costs and no moat.
I'd say 20% productivity growth with no increased salaries is a golden goose.
This comes from offloading 20% to AI as discussed, and still working same hours. It ignores transaction costs, incomplete rollout, but is a ceiling for long term growth in AI-friendly sectors.
I feel like we're talking past each other: a magic 20% productivity boost (or 100%, or 500%) for your company is worth nothing if your competitors all get the same magic dust.
I can be wrong without us talking past eachother. I re examine my assumptions on this every month or so. Right now, I see growth, but that might change soon.
Potable water is far more important than AI or iPads ever will be, but the world's most valuable water company only does about 5B/year in revenue: https://en.wikipedia.org/wiki/American_Water_Works