That's not what the author is arguing, he's simply stating that a main driver of this tech cycle - matching platforms for the sharing of private assets such as Uber and AirBnB - is running into problems, with appetite decreasing in public markets for companies that show (slowing) user growth with no clear path to profitability. A valid point.

What remains to be seen is the second claim - "the fantasy that AI and machine learning will generate trillions in profits"

While I tend to agree that ML/AI are generally deployed to save costs rather than increase revenue and create totally new categories of products, there are definitely still a lot of brownfield opportunities in legacy tech for data scientists to tackle and automate

This also ignores the huge opportunity for cloud service providers to take compute-as-a-service business from legacy on-prem/datacenter biz as well

But I would challenge you to think outside the sarcastic box and propose what you see as a few of the main technological drivers of productivity growth that are currently popping up in tech