HNHacker News
TopNewBestAskShowJobs

monk_grilla

431 karma · joined April 30, 2026

submissionscomments
monk_grilla··on Our position on open-weights models
The most American thing I've ever heard lmao
monk_grilla··on Our position on open-weights models
Yep. Sick of reading stuff that operates under the assumption that America is the faultless good guy and the only one to be trusted with the most dangerous and capable technology.

Among many other things, the Trump presidencies (and, to a lesser extent, the presidency between them) are examples of the the highest levels of leadership being totally incompetent, and have destroyed the above assumption for the rest of the world.

monk_grilla··on Our position on open-weights models
I think you're right. "Guardrails" as a concept has always struck me as a band-aid solution which any sufficiently motivated actor will circumvent by either bypassing them or using unrestricted, open-weight models.

In order to start securing and accepting our new reality we need to assume that capable, open-weight, unrestricted models will be widely available, and that their 3-6 month lag behind frontier proprietary models is just our forewarning of what attackers will soon be capable of. Trying to legislate against or control trade in such a valuable commodity is folly.

I also think that lag is going to shrink over time as the open-weight labs get more capable, acquire more hardware and the plateau starts to emerge.

monk_grilla··on Our position on open-weights models
> Someone who until yesterday did not seem bothered by his technology being possibly used to bomb elementary girls school in another country...

I might have missed something but wasn't the big story that Dario refused the Department of War's demand to use Anthropic's models for such purposes?

monk_grilla··on Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models
> Echo does not disclose its per-request routing decision because that policy is the product

I have not had need for a router product thus far so excuse my ignorance if this is standard, but how could I possibly use and improve a product built on a router like this if I am not permitted to see which model served my request? If I got a bad answer back in my LLM-powered app, do I really have no way of knowing which model was responsible?

monk_grilla··on Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models
I imagine that most small to medium sized businesses are on either individual plans or Teams plans. The vast majority of firms do not need more than 150 seats, and API rates are not sustainable for most.
monk_grilla··on Which streaming service was that on again?
Yeah this is truly crazy subtitle tech and it always blows my mind how they make it so seamless.
monk_grilla··on Businesses with ugly AI menu redesigns
> ... with software engineering going the way that it is. What lateral like passion could I transition to, as she did?

A question I ask myself increasingly frequently. Do let us know if you find any answer.

monk_grilla··on Terrence Tao's ChatGPT Conversation about the Jacobian Conjecture Counterexample
That's an interesting observation. I wonder if simply prompting for that in your personal system prompt might help?

"Be aware of 'xy problems': when the user asks x but really, they need the answer to y to further their understanding. ..."

The models are so capable now I think a lot of these real limitations can be solved by prompting, harnesses or fine-tuning.

monk_grilla··on Moonshine: Lets you stream games from your PC to any device running Moonlight
Do you leave your gaming rig switched on at home at all times?
monk_grilla··on The Voice of Google
I tried to verify that figure and I'm pretty sure you mean ~$28/user/year.

Instagram's global ad revenue ~$83.6B in 2025. [1]

Instagram passed 3 billion monthly active users in September 2025. [2]

$83.6B ÷ 3B ≈ $27.9 per user per year ≈ $2.30/month.

1: https://resourcera.com/data/social/instagram-revenue/

2: https://www.cnbc.com/2025/09/24/instagram-now-has-3-billion-...

monk_grilla··on Show HN: Hackney – Compare Uber, Lyft, Waymo, and Robotaxi Prices
Would love to give this a try once it is working in Australia. Our options are just DiDi and Uber.
monk_grilla··on Why write code in 2026
> not all of us did that. For many good reasons, too.

This isn't really the point of your comment, and for that I apologise, but I still haven't gotten used to speaking about what we "used to do" in the past tense, as if we are in a post-programming world...

We are already starting to speak of the habits we used to have (or not have), how things used to be done back in the day (1-2 years ago).

I don't think you're wrong to do so. I find myself saying it sometimes too. Just an observation about the weird world we find ourselves in.

monk_grilla··on Grok 4.5
Lol fair enough, I stand corrected.

It seems then that their objective is to superficially increase the diversity of results, to avoid bad PR, rather than actually neutralising harmful and untruthful biases of the input data.

monk_grilla··on Train sim created by just one person is being called the best ever made
Thanks for the recc, very cool extension!
monk_grilla··on GPT-5.6
This is the first I have herd of this benchmark. Can someone explain how it in any way indicates how close we are to "AGI"?

Replay of Sol attempting the game: https://arcprize.org/replay/83543d22-8e1e-439a-8809-129ff1d9...

It seems a weird and arbitrary challenge for a language model to be expected to perform. It also seems like there are some harness/visual issues even in the first few steps, where it states that it hasn't moved when it clearly has.

monk_grilla··on Grok 4.5
Do you think they deliberately trained the model to produce images of diverse Nazis?

It is clearly a byproduct of trying to correct an unaligned, bigoted model, and that is an example of overcorrection.

> Removing bias ends with truth, not these crazy wonky results.

Unfortunately there is an awful lot of untruth on the internet, if you hadn't noticed. This necessitates some correction through post-training.

monk_grilla··on Herdr: One terminal to rule them all
C-b + s or C-b + w for switching sessions is most convenient for me, but if I am running multiple sessions I would usually just open a new tab in my emulator (iTerm2) which can be clicked or hotkeyed between.

Anyway, I think multi-session management is a reasonable thing that this thing might improve over tmux, so fair enough.

monk_grilla··on Herdr: One terminal to rule them all
You can do this. In my current tmux session I have four windows each with a status indicator that shows up when that window is running a claude code session. It is a blue animated braille spinner when it is working, a red dot when it needs attention and a green tick when it is finished with its work.

Claude code exposes a bunch of hooks you can use to have tmux commands execute. This is invaluable if you are multi-tasking and need to see at a glance what needs attention.

monk_grilla··on Herdr: One terminal to rule them all
> wishing it was a little bit more mouse friendly

Can you explain what you struggle to do in tmux with the mouse? I'm using it as we speak and everything is clickable, panes are resizable, right clicking a windows give you a context menu... Not sure what else is missing?

monk_grilla··on Herdr: One terminal to rule them all
Tmux mouse support also has right-click context menus, clicking between tabs and panes, scrolling in panes, selecting text etc.
monk_grilla··on Claude Fable 5 Backlash Grows
Same. Not sure what it is but I have never hit the guardrails on Fable, and its knowledge and code is seriously impressive.
monk_grilla··on Has_not_been_viewed_much
This was my favourite of the ones I saw:

https://www.artic.edu/iiif/2/813984c9-f0a6-c340-5e89-f1c00af...

Really moving piece. Great idea on the part of the API devs!

monk_grilla··on Reddit will require you to log in to use old.reddit.com
How do you meaningfully use Reddit without an account? Do you really navigate to specific subreddits and read them one by one, rather than assembling a home feed of subs you have joined?

You never want to participate in discussion, or vote on comments and posts?

I don't really understand how you all were regularly browsing and using the site without an account.

monk_grilla··on Waveloop: What Fable left me
That generated video was eye opening for me. I've been using Opus in Claude Code for studying and at work, but it never occurred to me to use 3b1b's excellent python library for generating maths visualisations to let it generate such good graphical demonstrations.
monk_grilla··on Show HN: Live breath detection and biofeedback from a phone microphone
In the README it says

> Do I need an account to try it? Not for the core breathing biofeedback; that works without creating one. Some surrounding features in the app use an account; the part this repo is about does not.

I downloaded the app based on this, but you can’t use it at all without creating an account. Please review your AI’s output.

monk_grilla··on The advertising cartel coming to your web browser
Did you read it?
monk_grilla··on Mozilla to UK regulators: VPNs are essential privacy and security tools
That’s funny, I wonder if they might remove it since it is a common way for people to circumvent the ID requirement laws for certain sites.
monk_grilla··on The Emacsification of Software
I can’t be the only one for whom `vim README.md` is perfectly good. I’ve never considered the monospaced font a limitation, I prefer it. Coloured rendering works great and is all the visual aid I need to parse quickly.

I can see a table of contents being useful though. Perhaps if `:Toc` doesn’t exist yet, it should.

monk_grilla··on Wi is Fi: Understanding Wi-Fi 4/5/6/6E/7/8 (802.11 n/AC/ax/be/bn)
Anyone know of a similarly excellent resource for understanding wired networking? CAT specifications, how to pick high quality switches/routers etc.?
← PreviousPage 4 of 5Next →