OpenAI has 4,500 employees and a post-money valuation of $852 billion at its last funding round. You'd think they could get someone who specialises in writing MacOS apps to write their MacOS app.

They can probably hire them, and they'd get PIP'd for not jumping when management says jump.

Toxic corporate culture can go a long way to self-sabotage.

No. They want the AI to do it.

Hopefully at some point gravity pulls them back down and they reach the maturity and humility to realize that LLMs are just tools and even though they're good for many things, you can't do absolutely everything with them.

I think the "maturity and humility" phase of the LLM hype cycle is probably still 2-3 years into the future.

And that phase probably comes immediately after the "LLMs overinvestments have caused a massive global recession".

So have the AI build a hybrid Crux, Windows and web app with shared Rust base.

The genius play is to have an OSX expert write it with AI assistance, then let the press release say AI did it

I think the software out of the big AI labs: OpenAI's ChatGPT MacOSX app, Anthropic Claude Code - is a sign of the future to come.

Big, feature rich, built in quick iterations - but also, in particular if you look under the hood, of extremely poor quality if measured by traditional software engineering standards (code structure as exemplified by the leaked Claude Code source code, resource usage, "buggyness" etc.).

They absolutely do not have AI-heads with PhDs writing the apps and general infra. I've met a guy that worked at Anthropic, and he was "just an engineer", so exactly that: a top guy, but not specifically AI focused, doing regular stuff there but with infinite access to the top of the crop AI.

Now look at what software they make. It speaks for itself.