gchamonlive

I understand and aesthetics is pretty underestimated. If you want to get a message across, the form, the medium and the public needs to be accounted for. In this sense I totally agree.

But that kinda reminds me of the Chinese room thought experiment [1]. Is there anywhere in the latent space of the LLM a result that takes into account all these factors cited before? What kind of prompts or prompt orchestration would be necessary to achieve it? Would we still be able to tell the difference?

[1] https://en.wikipedia.org/wiki/Chinese_room

Can I take another stab at it? I take back what I said in my flagged comment. Could you elaborate on why the ethical points and the erosion of leadership in the article are banal and middle-of-the-road?

That's really interesting. I'll be more careful with RAG in the future then, thanks for the account

Separating emotion from logic without one dominating the other is a skill. That's Simone Weil's attention. You need to look at the world with curiosity and openness in order to filter out and assimilate only what you wish. See past what is for what in can be. I'm not saying the project is good or if it's aesthetic, but the functionality works and without LLMs this would never have seen the light of day. All these problems you say are easily solvable with something like Stitch, Impeccable and a tad of good taste from the part of the author. But for a MVP or a PoC? That's really more than enough.

I could take your comment and use it as a prompt and depending on the model quality it will have been an ex-bug

Great work! I'm doing something similar, but using upvoted submissions I did so I can rank posts better, filter out posts about AI but most importantly flag potentially interesting submissions in New that couldn't have made frontpage.

Wish you could read yourself. This is exactly like how horoscope works. "Born in October? Oh, you must be really insecure!"

Looks a lot alike Web UIs SOTA models produce, and you know what I'm really glad.

No more people doing crazy stuff with UX just because.

No more frontend dark magic only the most dedicated zealots can make sense of.

Democratised UI for all. And this might not even have had an AI assisted development, I'm just abusing this comment and going slightly off topic to say I'm happy for this tech, as someone that only dealt with backend and DevOps/SRE to be able to confidently create useful graphical applications for regular users and learning a ton about frontend architecture in the meantime.

Btw, a hall of fame for comments would also be interesting!

I really don't get passkeys and how they are supposed to be safer.

Currently I save all login tuples to Bitwarden and store OTP secrets onto Aegis. Could have been 1password and authy, it's irrelevant. The thing is, I only get pwned if both are compromised.

Now with ubiquitous passkeys in Bitwarden if someone has access to my vault unencrypted it's already endgame.

That's projection

Proof? Or are you treating a hunch as if it's fact?

Ah yes, I understand, but that's expected and actually desirable, it's preferable than having inconsistent project patterns across the project introduced by multiple LLMs. Imagine the mayhem.

It's important to do a first pass cleaning up and preparing code before unleashing agents onto it, similar to how Working Effectively with Legacy Code advocates approaching consolidated codebases. After that the agent will actually pick up the new pattern and start propagating that to the rest of the code.

Could have just moved on, but here we are now still wallowing in our own filth. The fact you realised and didn't just adds to the irony.

Writing a useless comment just to tell you: oh the irony, writing a garbage comment for calling out a garbage comment. I'm lost in the layers of slurry.

How about a bit of empathy. People don't have 10 years to wait for necessities, those vulnerable are suffering today

Is there such a thing as Reddit "unslop"? Thought it went without saying.

Can you even get pregnato

Sorry to break it to you but we are past end game since possibly the dot-com bubble.

"Cannot work well" and "cannot get started" are two different things. The whole of Latin America apart from Cuba is capitalist, for better or worse, regardless of how bad those countries keep their books.

The reason capitalism cannot work well is that it assumes endless resources and infinite growth, not because people don't like electronic transfers. Capitalism is failing in all countries and those that are not are already transitioning to some form or another or post consumerism.

Hyprland config break is usually syntax changes or reallocated features. Never had an instance where hyprland broke beyond repair or was hard to fix, even before LLMs were a thing. But yeah, its a 0.x software so if you aren't keen on fixing small stuff it's better to wait indeed. My point is that hyprland is easy to fix.

Not my experience. It does depend on the scale of the project, but for tens of thousands LOC, it's really good at taking instructions for architectural redesign.

LLM will take stuff at face value. If you tell it it's good, it'll gladly abide. Similarly if say it's crap it'll so the complete opposite.

Lots of land deals aren't dealt in money transfer either. It's sometimes cash, livestock transfer, other realstate or durable goods like cars or machinery.

Here in Brazil lots of people deal land by word of mouth. The country is just too big and there isn't enough state present to make every transaction official, so people shake on it, draft an unofficial document and actually living and developing the land over a long period of time is what consistitutes ownership.

The problem is that everything will potentially be under dispute, since anyone can claim they purchased a patch of land whose registry was missed in the restoration

I'd bet money the offline copy is far from up-to-date, given the state of the network. It still has potential to be mayhem

Nice. I'm currently slowly studying how Godot works and I'll next try to see if there is a usable workflow for designing assets in blender and using them in Godot.

How well do agents interact with blender? Do they modify files directly or through an API/mcp?

Rust does this automatically. It removes an entire class of errors from his backlog.

Even with the huge amount of "unsafe" rust currently in bun? https://news.ycombinator.com/item?id=48967630

All of which you had with only humans in the loop. Catalogue problems so they become technical debt and tackle them periodically. Seems to me like this is less of an AI problem and more of bad management.

It's useless to talk about models and harnesses without context and method. Depending on how you use the model and what the model is used for, experience may vary drastically. Also, different models with different harnesses require different approaches.

I've been using https://gitlab.com/gabriel.chamon/orisun which is my own simplified methodology, for coding web apps in python and elixir and have been very successful using qwen3.6 27b Q4 locally with help of larger models for architecture, so I get very suspicious when people talk how useless larger models are. They are either using it for a domain that models don't perform well or just not using it right.