SuboptimalEng on Hacker News

Ask HN: If AI wins, don't AI companies lose?

Let's play it out in the case of software engineers.

100 SWEs all running Claude Code. Okay, company is 10x more productive than before.

Now, company cuts 90% of SWEs because AI makes 10 engineers as productive as 100 engineers.

Yay, company saves 90% of engineering costs.

But now, 10 engineers run Claude Code. And now Claude revenue drops by 10x multiple?

So if AI gets good, it will reduce the very engineering headcount that AI companies depend on, right?

Right now many companies are in cost-cutting phase, so here is an AI thought experiment.

50% less white collar jobs -> 50% less productivity SaaS apps needed (slack, gmail, notion, jira, msft word, many more...)

Once company downsizes -> businesses that depend on said companies "growing in size" also lose.

AI doesn't need Slack, Jira, etc. That's just how we humans communicate. But AI can work directly with databases and store data more efficiently.

What now?

(FWIW, I do think AI is very useful. I use it daily to help me with engineering work. No arguments there.)