I can’t wait until it’s good enough to vibecode the next MongoDB.

Managing hundreds or thousands of containers across hundreds or thousands of k8s nodes has a lot of operational challenges.

Especially in-house on bare metal.

Yeah, totally agreed - there is still far too little negativity in comparison to what is going on.

I think you're missing the point of "AI winter". It's not about how good the products are now. It's about how quickly the products are improving and creating the potential for profit. That's what drives investment.

3 things we know about the AI revolution in 2025:

- LLMs are amazing, but they have reached a plateau. AGI is not within reach.

- LLM investment has sacrificed many hundreds of billions of dollars, much of it from the world's pension funds.

- There is no credible path to a high-margin LLM product. Margins will be razor-thin positive at best once the free trial of the century starts to run out of steam.

This all adds up to a rather nasty crunch.

The thing about winter, though, is that it's eventually followed by another summer.

There a single toggle setting for Apple Intelligence in MacOS (currently toggled as 'off' on my machine)

LLM-s also report that they enjoy my questions, in fact they tell me it's a good question literally every time I ask about their weird choices.

The license file states MIT and AGPL licences

Yes, Your Honor, I did convince this teenager to kill herself - but 150 people a year die from coconuts!

LLMs do take a lot of the joy out of it.

Intel should announce their new 40ni (40-nanoinch) process node next April Fools' Day.