> So long as their productivity was on par with the rest of the team there was no issue.

Emphasis added. And anyway, for most software dev in most shops it wasn't true; most development takes place in whatever IDE the group/organization standardized on for the task, to make sure everyone gets proper tooling and to make collaboration and information sharing easier. Think of all the Java enterprise software developed by legions of drones in the 2000s and 2010s. They all used Eclipse, because Eclipse is what they were given.

It's only with the emergence of whiny, persnickety Unix devs who refused to leave the comforting embrace of their editor of choice that shops in the internet/dotcom/startup tradition embraced a "use whatever tools you want" philosophy. They had uncharacteristically enormous leverage over the tech stack being deployed in such businesses and could force employers to make that concession. And anyway, what some of them could do with vi blew the boss's mind.

It is true that we don't have a whole lot of hard data from large organizations that show AI productivity improvements. But absence of evidence is not evidence of absence. Turns out, most large organizations just haven't adopted AI in the amount and ways that could make a big impact.

But we have enough anecdata from competent developers to suggest that the productivity gains are huge. So big, AI not only lets you do your normal tasks many times faster, it puts projects within reach that you would not have countenanced before because they were too complex or tedious to be worth the payoff.

So no. Refusing to use AI is just pure bloodymindedness at this point—like insisting on using a keypunch while everyone around you discovers the virtues of CRT terminals and timesharing. There were people like this even in the 1970s when IBM finally came around and made timesharing available in their mainframes. Those people either got up to speed or moved on to a different profession. They couldn't keep working the way they'd been working because the productivity expectations changed with the availability of new technology.

> It's only with the emergence of whiny, persnickety Unix devs who refused to leave the comforting embrace of their editor of choice that shops in the internet/dotcom/startup tradition embraced a "use whatever tools you want" philosophy. They had uncharacteristically enormous leverage over the tech stack being deployed in such businesses and could force employers to make that concession. And anyway, what some of them could do with vi blew the boss's mind.

They had enormous leverage because they were more productive then the drones who use whatever tools they are handed and lack the curiosity to use anything else. These breathless reports of increased productivity are constant, but why is there no evidence of that productivity increase otherwise? Why hasn't there been a surge of side-project video games on Steam? Why is github down so often despite Microsoft's commitment to AI?

The AI tools make it easier to do things that were already easy, but the minute your code gets interesting these tools are an absolute mess.