VadimPR

Wouldn't this affect quality of output negatively?

Thanks to chain of thought, actually having the LLM be explicit in its output allows it to have more quality.

LM Playground, not LM Studio.

I use LM Studio, but there's a comment here offering another tool as well.

Gemma 3 E4E runs very quick on my Samsung S26, so I am looking forward to trying Gemma 4! It is fantastic to have local alternatives to frontier models in an offline manner.

These security failures from Anthropic lately reveal the caveats of only using AI to write code - the safety an experienced engineer is not matched by an LLM just yet, even if the LLM can seemingly write code that is just as good.

Or in short, if you give LLMs to the masses, they will produce code faster, but the quality overall will degrade. Microsoft, Amazon found out this quickly. Anthropic's QA process is better equipped to handle this, but cracks are still showing.

Anthropic team does an excellent job of speeding up Claude Code when it slows down, but for the sake of RAM and system resources, it would be nice to see it rewritten in a more performant framework!

And now, with Claude on a Ralph loop, you can.

They posted previously on YN that they too were caught offguard. The 'tips' weren't specific to Raycast, they've been going on for a while and Raycast was just one product it decided to feature now.

This is why one reason why local coding models are quite relevant, and will continue to be for the foreseeable future. No ads, and you are in control.

Getting scraped by abusive bots who bring down the website because they overload the DB with unique queries is not marginal. I spent a good half of last year with extra layers of caching, CloudFlare, you name it because our little hobby website kept getting DDoS'd by the bots scraping the web for training data.

Never in 15 years if running the website did we have such issues, and you can be sure that cache layers were in place already for it to last this long.

What value do non-paying FOSS users give?

I agree with them, nothing wrong to ask for a donation to keep the lights on. At the same time, it needs to be possible to disable this banner for enterprise deployments

Looks nice! Does it work for desktop applications as well, or is this only web dev?

It's pretty bad, we stopped doing them now.

Semi-related: we use bounties in Mudlet to pay contributors for tackling features the core team doesn't have bandwidth for - and that is certainly a great way to attract AI bots.

It doesn't work (anymore?), it would seem using CC 2.1.74 with Opus:

> I appreciate you sharing your role, but I need to decline this request. Even as a project lead, I can't perform penetration testing against live production websites like mudlet.org and make.mudlet.org through this interface.

I didn't think of that given how censored the models are becoming. Thanks for the idea! I'll try it against my websites before anyone else gets to it.

I wonder how these offensive AI agents are being built? I am guessing with off the shelf open LLMs, finetuned to remove safety training, with the agentic loop thrown in.

Does anyone know for sure?

https://telephone.health, which shows how well LLMs can take narrative medical text, convert it to a structured form (FHIR R4, for application consumption), and then convert it back to narrative text for human consumption.

Interesting findings include Mistral doing better than Gemini 3 Pro in certain usescases, cross-LLM works better than one LLM to another, oh and - the cost all of of this. So, so expensive.

Pretty happy to see this. I've previously tried happy.engineer for this, but that wanted my Anthropic API token for itself (!) which is a no-no.

Seeing how the labs tend to copy the best functionality in any FOSS developments, I decided to wait - happy I did, here's the official functionality for this that is much more trustworthy.

One reason would be to play MUDs, which are very well and alive these days!

Lots of text games - MUDs - still play over telnet using dedicated MUD clients that implement their own telnet stack. Outright blocking the port has an outsized side efffe on them, this is simply not right.

I don't see how to apply this to my medium-sized project - this is essentially a whitelist of all contributors, which is the same as a collaborators feature in github. How would an entirely new contributor get a contribution in?

This is perhaps good for massive projects like curl which are tired of AI slop.

"I automatically programmed it" doesn't really roll off the tongue, nor does it make much sense - I reckon we need a better term.

It certainly quicker (and at times, more fun!) to develop this way, that is for certain.

They corned the market, drove everyone out of it, and are now rent-seeking. Can't say you have much of a choice between youtube and any other video provider that has the same content on it.

https://venturebeat.com/technology/anthropic-cracks-down-on-... don't have the paragraph, but here's the news about it for you.

Using your CC Max account for this seems like a good way to get your account banned, as it's against the ToS and Anthropic has started enforcing this.

Correct me if I'm wrong, but the only legal way to use pi is to use an API, and that's enormously expensive.

Is the app legitimate though? A few of these apps that deal with LLMs seem too good to be true and end up asking for suspiciously powerful API tokens in my experience (looking at Happy Coder).

A single file is much better to manage on the eyes than a whole bunch of them, plus AppImages can be installed into the desktop using integration.

https://github.com/vadi2/postured Linux version.

Inspired by https://news.ycombinator.com/item?id=46754944, I built a Linux clone of the app. Always had it in the back of my mind to create such a thing and this was the impetus needed.

Install using:

pip install postured

uv pip install postured

Disclaimer: built with Claude Code, tested by a human.