Calazon

Exhaustive brute force prompting is completely unfeasible. The number of potential prompts is impossibly large.

Psychologists and therapists have different specialties too, for mental differences. This is generally considered a good thing.

Hopefully he would be using the LLM as an enhanced search engine that can point him to relevant authoritative sources that he can use to fact-check its output. I have done that in the past to some effect.

With that diversity of preferences, some organizations might also be willing and able to do rigorous testing of the updates that are most important to them.

It seems like a helpful efficiency to spread out the testing burden (both deliberate testing and just updating and running into unexpected issues). If everyone updated everything immediately, everyone would be impacted by the same problems at the same time, which seems suboptimal.

If you like Catholicism but struggle with the papacy, have you considered Eastern Orthodoxy?

Do you even use two hands for shortcuts like copy and paste?

> Sometimes the docs straight up lie, and it takes 5 minutes to figure that out. Should they also be ashamed?

Yes.

I feel like we do that pretty regularly in the US.

If 1% of the people in each specialization are advancers, and you add up all the specializations together, then 1% of the total number of people are advancers.

Even this assumes that everyone has a specialization in which 1% of people contribute to the sum of human knowledge. I would probably challenge that. There are a lot of people in the world who do not do knowledge-oriented work at all.

"Vibecoding" is about how you use AI tools, rather than who you are.

If you're asking the AI to generate large amounts of code that you don't really look at, you're definitely vibecoding. If you are mainly writing the code yourself with some assistance from AI tools, you are not vibecoding.

Naturally it's fuzzy in the middle. And for people without software development skills, vibecoding is the only option they have.

Most documentation is documenting things that somebody already wrote down in a different form.

The quality of AI-made documentation may be poor, but calling it 0% is just silly.

If you aren't any closer to the goal after the step than you were before it, you didn't take a step towards the goal.

I doubt Russia would invade the EU anytime soon under current conditions, even if they manage to annex all of Ukraine with no further losses.

Much more likely they invade another non-EU, non-NATO country like Moldova.

I recently used Preact and HTM for a small side project, for the JSX-like syntax without a build step.