The elephant in the room

Colton Voege has penned an excellent debunk of the 10x AI engineer. Their account of rapidly diminishing returns mirrors my experience with agentic coding. Yes, it can create 10x the code that you or I could create in the same time period, but it certainly can’t create 10x the value or impact.

Through lunchtime chats, off-hand comments and a growing number of DM’s, it’s becoming clear that, although the AI propaganda is loud, it’s not landing with a great deal of engineers on the ground. Many I’ve spoken to appreciate its use within a niche set of circumstances, but I could count the number who’ve actually drunk the koolaid on one hand.

And yet, within a tech company, it’s taboo to voice any form of healthy scepticism.

If you are an engineer and your boss asks you:

Hey, you’re getting 10x the productivity thanks to AI, just like all the other engineers, right?

You are strongly incentivized to say yes. And when every other engineer also says yes for the same reason, that CEO isn’t lying, they are just relaying what they heard.

No, AI is not Making Engineers 10x as Productive, Colton Voege

But when the rhetoric is “your very livelihood is on the line if you don’t subscribe to this brave new world”, we really should be able to discuss this in the open.

Regardless of whether you believe AI is coming for your job or not, navigating any existential threat alone is exhausting. It’s time to re-read You Deserve a Tech Union and plan a forum for these discussions…