So I find this to be interesting. Given that the author explicitly stated that he’s not very familiar with the industry, their viewpoint is of one who’s external to the whole bubble.

A lot of the article seems to run under the purported productivity improvements that companies make, especially considering what they mention about LLMs generating “world-class” code. Anyone who uses it for more serious projects knows there is a lot of nuance that’s not really well-captured by code-gen models (and I can see some of the HN comments giving similar experiences)

I see this as a problematic disconnect. Those on the outside have no idea of the reality on the ground, which leads them to have cautious optimism. I’m of the opinion that such opinions would quickly sour if they see the actual productivity loss.

To which I say, it’s going to take a noticeable loss of productivity for companies to recognize that they’re paying extra to get less, and i wonder if this would be the marker for this bubble popping.