HNHacker News
TopNewBestAskShowJobs

trenchgun

1,050 karma · joined October 28, 2019

submissionscomments
trenchgun··on Do I belong in tech anymore?
CEO observes performative work, and his inference will be that means more people need to be fired. Let only the AI native, customer obsessed 10x engineers(/ AI swarm managers) remain.
trenchgun··on Category Theory Illustrated – Orders
This is LLM written comment
trenchgun··on Category Theory Illustrated – Orders
How is this useful?
trenchgun··on Copilot is 'for entertainment purposes only', per Microsoft's terms of use
Which of the 78 copilots? https://teybannerman.github.io/strategy/2026/03/31/how-many-...
trenchgun··on Caveman: Why use many token when few token do trick
You’re a literature cycle behind. ‘Middle-layer shared representations exist’ is the observed phenomenon; ‘why exactly they form’ is the theory.

You are also confusing ‘mechanistic explanation still incomplete’ with ‘empirical phenomenon unestablished.’ Those are not the same thing.

PS. Em dash? So you are some LLM bot trying to bait mine HN for reasoning traces? :D

trenchgun··on Claude Code runs Git reset –hard origin/main against project repo every 10 mins
Yes but this is probabilistic. Skill, documentation etc help by giving it the information it needs. You are then in the more correct probability distribution. Fine for docs, tips etc, but not good enough for mandatory things.
trenchgun··on The first 40 months of the AI era
Yes this + also APIs! Due to LLMs needing APIs suddenly things that were strictly behind GUI can easily opened up for programmatic use. Double edged sword, due to infosec etc concerns, but sure is nice for personal productivity automation.
trenchgun··on Show HN: Cq – Stack Overflow for AI coding agents
Do you run security review by agents over this?
trenchgun··on Regex Blaster
Claude default frontend skills suck :D
trenchgun··on Wayland set the Linux Desktop back by 10 years?
Why is it better than native x11?
trenchgun··on Leanstral: Open-source agent for trustworthy coding and formal proof engineering
>Such is very difficult at the moment

What do you mean? It's a nice and simple language. Way easier to get started than OCaml or Haskell for example. And LLMs write programs in Lean4 with ease as well. Only issue is that there are not as many libraries (for software, for math proofs there is plenty).

But for example I worked with Claude Code and implemented a shell + most of unix coreutils in like a couple of hours. Claude did some simple proofs as well, but that part is obvs harder. But when the program is already in Lean4, you can start moving up the verification ladder up piece by piece.

trenchgun··on 1M context is now generally available for Opus 4.6 and Sonnet 4.6
LLMsdo great with Rust though
trenchgun··on Show HN: LocalGPT – A local-first AI assistant in Rust with persistent memory
You could have a multi agent harness that constraints each agent role with only the needed capabilities. If the agent reads untrusted input, it can only run read only tools and communicate to to use. Or maybe have all the code running goin on a sandbox, and then if needed, user can make the important decision of effecting the real world.
trenchgun··on Monty: A minimal, secure Python interpreter written in Rust for use by AI
Python has uv, ruff, ty
trenchgun··on Claude is a space to think
You need a mind mapping tool there
trenchgun··on Julia
Baroque
trenchgun··on Auto-compact not triggering on Claude.ai despite being marked as fixed
Claude Code is mostly vibe coded app, as is Claude. I guess this should have been expected.
trenchgun··on Are arrays functions?
Arrays/maps/lists are extensionally defined functions, where as functions/TLA+ operations are intensionally defined functions
trenchgun··on Ask HN: Should "I asked $AI, and it said" replies be forbidden in HN guidelines?
Permaban from first strike
trenchgun··on DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning [pdf]
LLMs are already a powerful tool for serious math researchers, just not at the level of "fire and forget", where they would completely replace mathematicians.
trenchgun··on IDEmacs: A Visual Studio Code clone for Emacs
What on earth are you talking about?
trenchgun··on Ruby Blocks
> The biggest thing holding Ruby back is lack of gradual typing imo.

There is: https://sorbet.org/

trenchgun··on Kiro: A new agentic IDE
Kiro in Finnish means "a curse".
trenchgun··on Successful people set constraints rather than chasing goals
Constraints create games.
trenchgun··on GitHub issues is almost the best notebook in the world
Well... not everybody uses github, and not every task in Jira is about code in version control
trenchgun··on Clojuring the web application stack: Meditation One
Also: NASA used to use Common Lisp before
trenchgun··on The great displacement is already well underway?
Yep.

Some context from the blog post: > I turned to service apps this winter: doordash, instacart, uber eats. Their signup systems were incompatible with my full, legal, one-letter last name, and it took about 50 hours on the phone with doordash support in Malaysia and the background check provider in India to eventually get cleared to drive them. I was not able to get through on the other apps.

For sure the impact is not just limited to service apps.

trenchgun··on We fell out of love with Next.js and back in love with Ruby on Rails
> Of course, Figma‑ or Gmail‑class apps still benefit from heavy client logic, so the emerging pattern is “HTML by default, JS only where it buys you something.” Think islands, not full SPAs.

Figma is written in C++ to webasm.

trenchgun··on One Million Chessboards
This would be fun!
trenchgun··on Microsoft subtracts C/C++ extension from VS Code forks
Abandon VSCodium, Return to Emacs
← PreviousPage 2 of 18Next →