There is nothing like the scorn of a developer who bought into something and then realizes that they were very wrong.
"Where's the Shovelware? Why AI Coding Claims Don't Add Up": https://mikelovesrobots.substack.com/p/wheres-the-shovelware-why-ai-coding
Four months later and the data coming back is saying a very similar story, there is no productivity goldmine here:
> "if you're confused how other people can be so productive, you're not broken. The data backs up what you're experiencing."
... that really spoke to me.
I was rereading "Peopleware" the other day and realized where the 10x myth came from:
https://hachyderm.io/@shafik/113749153071956255
it was such a basic misinterpretation of the data.
Why as an industry are we sooo bad at this? Discrete Math is one of the foundations of Computer Science courses. One would think we could get our act together on understand statistics.
6/x The 10x dev myth, and the reality https://hachyderm.io/@shafik/113746406038257681
@shafik reminds me of this article https://nibblestew.blogspot.com/2025/08/lets-properly-analyze-ai-article-for.html It claims that, in the Soviet Union, percentage comparisons were often made with 1913, the worst year for Russia as a whole. Very convenient.
"So if you encounter a claim that "car manufacturing was up 3700%" some year in 1980s Soviet Union, now you know what that actually meant."
Martin
Shafik Yaghmour
LR
Oliver Reyse