[mold_aid]

I think this really gets at it: people are so terrified of not knowing what to do, of not knowing whether their solution is "good," that they'll pay a monthly fee for a machine to tell them it's ok, ironically bypassing human judgment in the end. Drudgery or judgery, those are the two task contexts in which AI products* excel.

* It's lovely to have the opportunity to disagree with both Gruber and the "the whole thing smacks of politics" HN commentariat, pulled daily between "it's just a tool, like a hammer, which also kills people, stay with me here" and "AI puts an expert in your pocket; soon, the expert will live in your eyes"

[lxgr]

I’m not terrified of learning how to design a good website, I would simply not take up a design-heavy project, as I don’t enjoy it at all.

I do like certain other aspects of projects like that, which is why it’s great to have LLMs to collaborate with on it.

> they'll pay a monthly fee for a machine to tell them it's ok

Why would I outsource that decision to an LLM? I can look at the result myself and decide if I like it.