HNHacker News
TopNewBestAskShowJobs

josefrichter

1,369 karma · joined March 21, 2011

Product designer at josefrichter.design

contact josef.richter at me.com

submissionscomments
josefrichter··on Don't fall into the anti-AI hype
The "AI stealing jobs" is just a new-age Luddism. Yes, of course it will be painful for many, but at larger scale, even those who lose the jobs should benefit overall.

"your ability to create a mental representation of the problem to communicate to the LLM" – this is the tipping point imho. So far, you need to be good at this. That's why senior jobs are not affected yet. The question is for how long. We are probably just months away from the time when LLMs (or other form of AI) will be better at creating better "mental representation", better abstractions and better solutions, than most humans in most cases, including those in senior positions. And that will spill over to other non-dev jobs too.

josefrichter··on The Napoleon Technique: Postponing things to increase productivity
Didn't he basically observe the second law of thermodynamics?

The Second Law says an imbalance (like heat in one corner of a room) will "solve itself" by spreading out until the room is balanced. The Napoleon Technique assumes that a "social imbalance" (a crisis, a frantic email, or a minor conflict) will often "cool down" or reach a state of balance on its own if you simply wait.

In my part of the world, we call this technique "let it rot out" :-)

josefrichter··on Codex vs. Claude Code (today)
I think a big part is which model seems to work better with your language/stack. My language is Elixir, which is somewhat niche, and only Claude has been able to produce usable Elixir code so far. None of the other things mentioned in the article mattered, because of this. I wonder if others have this experience that some models just struggle with some languages/stacks?
josefrichter··on 'Ghost jobs' are on the rise – and so are calls to ban them
That's in ideal world. In real world there's maybe 1% chance your cover letter will reach a human being.
josefrichter··on 'Ghost jobs' are on the rise – and so are calls to ban them
Cover letters are dead imho. Even before AI came to play.
josefrichter··on 'Ghost jobs' are on the rise – and so are calls to ban them
Oh I am pretty sure it’s bad. I am actually quite shocked someone would ruminate wherever or not it’s bad.
josefrichter··on 'Ghost jobs' are on the rise – and so are calls to ban them
I believe it actually is illegal, via a set of more general rules on data collection, on what constitutes a fraud, etc. May not be spelled out exactly like this specific use case, but still very likely covered by law. Just difficult to prove.
josefrichter··on Italy's longest-serving barista reflects on six decades behind the counter
This lady keeps popping up on Hacker News every 3 months :D God bless her!
josefrichter··on X Just Accidentally Exposed a Covert Influence Network Targeting Americans
The scandal is he most likely knew the election is being manipulated via his platform, but didn’t say anything, because it was in “his” candidate’s favor.
josefrichter··on X Just Accidentally Exposed a Covert Influence Network Targeting Americans
So... Is Musk really trying to say he didn't know this before election? This has to be investigated, as it has implications far beyond Musk. This is basically a global scandal.
josefrichter··on Racket v9.0
For all lisp lovers tackling concurrency and parallelism, a reminder there's also LFE - a lisp for Erlang VM. Which happens to have solved all these issues decades ago. LFE was created by one of the Erlang co-creators. https://lfe.io
josefrichter··on Erlang Meets Idris: Cure Programming Language
How does this compare to Gleam, in terms of goals, features, etc.?
josefrichter··on No Socials November
Damn, I might try this, just to get a better perspective. Does HN fall within “socials” though? :-)
josefrichter··on Code like a surgeon
Interestingly, I used the exact same metaphor the other day. Suggesting that tools like GitHub Speckit should have /wrap-up command:

> And then maybe "/wrap-up" to tie all the untied knots once you're sufficiently happy. Kinda like surgeon stepping aside after the core part of the operation.

josefrichter··on The death of thread per core
Isn't this what Erlang/Elixir BEAM is all about?
josefrichter··on Gleam OTP – Fault Tolerant Multicore Programs with Actors
Come and have a chat at elixirforum.com. Plenty of folks write serious stuff entirely/predominantly in Erlang/Elixir & BEAM and will be happy to answer your questions.
josefrichter··on Claude Skills are awesome, maybe a bigger deal than MCP
From the docs:

"Skills work through progressive disclosure—Claude determines which Skills are relevant and loads the information it needs to complete that task, helping to prevent context window overload."

So yeah, I guess you're right. Instead of one humongous AGENTS.md, just packaging small relevant pieces together with simple tools.

josefrichter··on Understanding Spec-Driven-Development: Kiro, Spec-Kit, and Tessl
yeah you need to read through its templates and "source code" to understand what it does - which is not necessarily a bad thing for this type of project.
josefrichter··on Understanding Spec-Driven-Development: Kiro, Spec-Kit, and Tessl
I like the part that uses custom slash commands as way to wrap your input into some well-structured prompt template for given type of task. I like the part that also injects relevant pieces of "broken down AGENTS.md" as I see it.

I don't like the part that tries to leave no knot untied, which creates that sledgehammer for cracking a nut, as mentioned in the article. But I am sure it's easy to add another custom slash command like "/experiment" or "/stub" that would bring those context management benefits without the bloat, in situations when you don't know yet what and how you want to build something.

And then maybe "/wrap-up" to tie all the untied knots once you're sufficiently happy. Kinda like surgeon stepping aside after the core part of the operation.

josefrichter··on Ask HN: Abandoned/dead projects you think died before their time and why?
Concorde :-)
josefrichter··on Ask HN: Abandoned/dead projects you think died before their time and why?
It was actually fantastic even for creating websites. To think that 20 years later we still don't have tools to make similar stuff with similar ease is mindblowing.
josefrichter··on Examples are the best documentation
Elixir world makes this easy:

We advise developers to include examples in their documentation, often under their own ## Examples heading. To ensure examples do not get out of date, Elixir's test framework (ExUnit) provides a feature called doctests that allows developers to test the examples in their documentation.

Doctests work by parsing out code samples starting with iex> from the documentation. You can read more about them at ExUnit.DocTest.

josefrichter··on After nine years of grinding, Replit found its market. Can it keep it?
I don't think they can keep it. What they have has already been replicated by the big guys, with just maybe a few months delay. It seems like they're usually going in the right direction, even ahead of others, but it's usually not any massive breakthrough, they're just being good at executing quickly the natural next steps.
josefrichter··on A simple habit that saves my evenings
“Diffused thinking” is the official term for the magic behind Pomodoro technique, I believe.
josefrichter··on Comprehension debt: A ticking time bomb of LLM-generated code
I guess this is just another definition of vibe coding. You're deliberately creating code that you don't fully understand. This has always existed, but LLMs greatly amplify it.
josefrichter··on Learning Persian with Anki, ChatGPT and YouTube
I share your skepticism. The approach you mention seems to be not only the most efficient, but maybe actually the only one that works. This is probably more about neuroscience than personal preferences.

The approach presented in the post is not necessarily on conflict with that, but it seems like extremely laborious preparation stage. And despite expectations, learning the different alphabet is usually unnecessary until fairly late into the learning process.

josefrichter··on Learning Languages with the Help of Algorithms
I am fiddling with some language learning utilities myself. Can anyone recommend some relatively simple ways of tracking users' knowledge of a given language? Something like having a sorted frequency list of words/phrases/concepts, and tracking how many times each word has been seen vs used correctly vs used incorrectly, etc.?

I believe this does not have to be perfect, simplicity is preferred. But it should be just enough for an LLM to take a glimpse and estimate users' level in given language.

josefrichter··on Claude Code: Now in Beta in Zed
In Elixir/Rails (and newly React) world you can connect Clade Desktop to Tidewave, which gives you pretty close integration with codebase and CLI tools. That makes the experience somewhat similar to running Claude Code in editor's terminal, or now directly connected in Zed.

I think basically all these tools are trying to integrate all the stuff together and find the best way to do that, so that you don't need to be copy-pasting stuff around, jumping between tools, etc.

josefrichter··on Claude Code: Now in Beta in Zed
Does it? Or do you mean the integration when you run Claude Code from the terminal opened in VSCode? (which is already pretty good)
josefrichter··on A staff engineer's journey with Claude Code
I'm almost sure that we all ended up at the same set of rules and steps how to get the best out of Claude - mine are almost identical, others' I know as well :-)
← PreviousPage 4 of 19Next →