Philpax

And with the right kind of engineering, that could be automated too. I don't think the results would be as good as they are now - Vedal is bringing his own perspective into it - but once you have…

If you enjoyed it, does it matter?

I don't think there's a single human on or outside of this planet that can meet that requirement, but Claude has been pretty good to me. It's certainly a much better starting point than…

I don't know how to adapt those settings to the corresponding module - there are often differences in naming and hierarchy conventions - and there are other NixOS-specific considerations with reg…

Hate to be that guy, but which LLM was doing the generation? GPT-4 Turbo / Claude 3.x have not really let me down in generating ffmpeg commands - especially for basic requests - with most of thei…

My experience has been in complete agreement with yours: I love the theory, but the practice is so, so painful. And yes, I also had to settle for your NVIDIA fix. I suspect I would have had a marginal…

I wouldn't be surprised if most LLMs already knew how to do most of these out of the box.

Seems like all of those points about the landscape are exactly why you'd want to work on a personal blog. You can write something that's representative of you . As for the how: same as al…

Definition 3: https://www.dictionary.com/browse/violence

If a CEO of the largest health insurer is "a cog in the wheel", who isn't?

Ok: https://fortune.com/2024/12/05/unitedhealthcare-ceo-brian-th... …

I don't see how that's a technological problem? It's certainly a social problem (your agent's provider might have a deal with airlines / hotels to gouge you), but there's…

That's not what they said, and you're being purposely obtuse. They were saying that Llama 3.3, an open model, approaches usefulness, in comparison to o1, a closed model, which is already use…

It's about the journey, not the destination.

The Bitter Lesson states that you can overcome the weakness of your current model by baking priors in (i.e. specific traits about the problem, as is done here), but you will get better long-term resul…

> When there is such a rich database of manual pages and q/a about these tools, I tend to blame the user rather than the tool when I hear it called "too complex". Extensive documenta…

> These bugs persisted while they were introducing AI features that I really don't care (it's a terminal, why would we need AI???). Many terminal programs, especially older ones, are know…

Do you know who Simon is?

I'm no big fan of OpenAI myself, but what he was whistleblowing over (OpenAI training on copyrighted material) is not really a secret, nor is it the kind of knowledge an organisation would kill o…

Compilers can detect errors in the grammar, but they cannot infer what your desired intent was. Even the best compilers in the diagnostics business (rustc, etc) aren't mind-readers. A LLM isn…

Losing potentially good talent because they don't know how or where to look for mistakes yet is foolhardy. I'm happy for them to throw in the towel if the field is truly not for them, but I …

The use of a LLM in this case is to show them where the problem is so that they can continue on. They can't develop an understanding of the algorithm they're studying if they can't get …

https://en.wikipedia.org/wiki/Self-balancing_scooter Not the most popular, especially these days, but they are very much descended from Segways and have their own fans.…

I suspect they're referring to the Infowars kind of supplements, not vitamin supplements.

I don't think we need to pretend Python package management is good to contrast C++, they're both awful for different reasons. At least there's maybe a way out for Python (uv / the …

I'd suggest you level up your reading comprehension before suggesting the parent poster was in any way offended or in need of therapy.