Viewing profile — geraneum

I’ve read LLM outputs on a piece of code or topic that I already understand or hand written before, and lately, it’s so confusing sometimes that I need to reread a couple of times …

We, developers/tech-employees (safe to assume since we’re on HN) are the “entire class of workers” that the article is talking about. The premise is losing faith in the future of o…

Remember, our misery is some potential investor’s financial prospect. We read these articles with different eyes. There are people who are the real audience and salivate when they …

The author has discovered the red tape.

Blimey, it’s a mystery why they don’t let me take my jigsaw to the “community of hand saw enthusiasts”.

> observe that the agent day after day do handle user input safely How do you observe the issues that aren’t apparent via a GUI? Do you notice the circular logic in your reasoning?…

> I don't review code that either works or doesn't - most HTML and CSS layout code for example. There I test it on desktop and mobile and commit it if it works. Good example of wha…

> It's novel because previous rounds of automation were about automating specific tasks or well-scoped functions. Evidently, this has not really changed with LLMs and coding agents…

Does the code get reviewed? How do you deal with increased amount of code that may need to be looked at?

Unfortunately people sometimes get defensive against this take. But I think treating the LLM as you described can make you a better LLM user and help get better output. It helps un…

Then ask it to fix it. When “fixed”, ask the same question again and you’ll get a similar response again!

I’m terribly sorry on their behalf. I hope the expression of their experience has not hurt Claude’s feelings (IPO valuation). Won’t happen again.

> Never write READMEs, docstrings, or comments. I will write those myself later. And yes, I really mean this. The precise and rigorous practice of “engineering” in 2026.

That pesky “scientific method” bothers these corps sometimes, so one can play fast and loose with numbers and conclusions for all sort of reasons. No need to worry about reproducib…

This is what you get when you prompt claude to avoid –

You’re absolutely right to push back. It was indeed a school. I’ll save in my memory for the future.

Stata is better in statistics than we are. Why doesn’t that produce guesses?

> Anthropic has never advocated for a ban on open-weights models. If you wonder why this is written as an opening to a list of reasons that advocate for banning the open weight mod…

How does the AI guess?

Even on a less dangerous level, repeated mistakes can be detrimental to your business. Imagine a typical ecommerce app for selling anything. Mess with people’s orders and money and…

It might be because it was going out of its way before and had too much of a blast radius, and now they could have changed the RLHF (or other tricks in their sleeve) to get what yo…

I read this as the model being less steerable. I was bitten by this just today where I had a local Postgres instance running, and prompted opus 5 to run a server against it, but fo…

What better way to spend token. I’m amazed that people talk about this as it’s a good thing!

Why not make the agents make what you want with marimo?

> It largely works and it's a massive business success. The engineer is suggesting that it could be done cheaper and maybe with better outcome. Ironically, this is a classic busine…