bonoboTP

31 ms·

751.
by bonoboTP 9mo ago
Most papers are badly written. https://en.wikipedia.org/wiki/Sturgeon%27s_law
752.
by bonoboTP 9mo ago
If you dress down in a context where formal attire is expected, it's a signal. What it signals depends on what happens. If you're shunned and avoided, then you're just a loser or a hobo. If you're clearly valued, listene
753.
by bonoboTP 9mo ago
You can call it whatever you want but people make inferences. Also there is no bright line between intentional and unintentional signaling. The brain is capable of hiding plenty of stuff from its own other parts. See the book "The elep
754.
by bonoboTP 9mo ago
And that's only bad if it's illusory or fake. This reaction evolved because it's adaptive. In slot machines the brain is tricked to believe there is some strategy or method to crack and the reward signals make the addict feel
755.
by bonoboTP 9mo ago
A signal is a two way street. It remains a signal even if the signaler is oblivious to it but the observers still draw conclusions.
756.
by bonoboTP 9mo ago
People are always impressed by how formal and informal tone and relative status is encoded in East Asian languages and how English doesn't have this and is supposedly egalitarian. Here's an example to show how it does exist also i
757.
by bonoboTP 9mo ago
Yes, overthinking this is also possible. I've had bosses who type correctly capitalized, with punctuation and paragraphs, and it's simply their style, not much else to read into it. But sometimes it can indicate a certain pedantic
758.
by bonoboTP 9mo ago
And of course you can, at latest after one or two sentences.
759.
by bonoboTP 9mo ago
You can't really opt out, just choose better suited minigames. Generally when you don't (have to) care, you either have to back that up with some other accumulated reputation/value, or sacrifice some things. Like you can opt
760.
by bonoboTP 9mo ago
They can try it and sometimes it works, but generally it's hard to imitate well. You have to not give a fuck about the right things. The imitators who just don't give a fuck about anything will stumble on something genuinely impor
761.
by bonoboTP 9mo ago
Not necessarily but it carries less weight than pre-LLMS. Obviously it's just a heuristic and not the whole story and telltale AI signs are not purely about good spelling and grammar. But I just appreciate some natural, human texture i
762.
by bonoboTP 9mo ago
Not having to care is often part of the countersignaling. An honest signal doesn't always take effort. In fact it's the tryhard imitators that have to expend effort emulating this. The real deal is effortless and comes naturally.
763.
by bonoboTP 9mo ago
On the positive side of this, research papers by competent people read very clearly with readable sentences, while those who are afraid that their content doesn't quite cut it, litter it with jargon, long complicated sentences, hoping
764.
by bonoboTP 9mo ago
Exactly. If Codex is really as good, it should have no problem porting any settings or config from the Claude setup. (and I do believe it wouldn't have much of a problem)
765.
by bonoboTP 9mo ago
Kai Lentit is great!
766.
by bonoboTP 9mo ago
One use I have for seeing what exactly it is doing is to press Esc quick when I see it's confused and starts searching for some info that eg got compacted away, often going on a big quest like searching an entire large directory tree e
767.
by bonoboTP 9mo ago
Do you think this is a fundamentally unbridge-able limitation of LLMs? Do you know where we were just a year ago? Can you imagine that this will get better with upcoming releases? It's like when Gary Marcus was confidently stating that
768.
by bonoboTP 9mo ago
I think it's more classic enshittification. Currently, as a percentage, still not many devs use it. In a few months or 1-2 years all these products will start to cater to the median developer and start to get dumbed down.
769.
by bonoboTP 9mo ago
It's unclear where the car is currently from your phrasing. If you add that the car is in your garage, it says you'll need to drive to get the car into the wash.
770.
by bonoboTP 9mo ago
Not in my experience so far. But if it becomes a problem, I'll reconsider.
771.
by bonoboTP 9mo ago
I personally am a lot less stressed. It helped my mood a lot over the last couple of months. Less worries about forgetting things, about missing problems, about getting started, about planning and prioritizing in solo work. Much less of the
772.
by bonoboTP 9mo ago
Local models exist and the knowledge required for training them is widely available in free classes and many open projects. Yes, the hardware is expensive, but that's just how it is if you want frontier capability. You also couldn&#x27
773.
by bonoboTP 9mo ago
I find working with LLMs much more fun and frictionless comprated to the drudgery of boring glue code or tracking down nongeneralizable version-specific workarounds in github issues etc. Coding LLMs let you focus on the domain of you actual
774.
by bonoboTP 9mo ago
Commercial ventures already had to care exactly to the extent that they are financially motivated by competition forces and by regulation. In my experience coding agents are actually better at doing the final polish and plugging in gaps tha
775.
by bonoboTP 9mo ago
Hacker news is a bit different from Reddit or other social media. This is a good summary, especially the section titled "Comments". https://news.ycombinator.com/newsguidelines.html Edit: I notice I'm talking
776.
by bonoboTP 9mo ago
It's the top comment. But I'll disengage as this gets too meta. I'd encourage talking about the substance.
777.
by bonoboTP 9mo ago
This comment has even lower nutritional value. It's just a "dislike" with more words. You could have offered your counterarguments or if you're too tired of it but still feel you need to be heard, you could have linked t
778.
by bonoboTP 9mo ago
Not sure if I get it, but it seems to me that this is not for "producing code" eg for your projects or doing things on your computer but essentially for supplementing its own thinking process. It runs this python code to count how
779.
by bonoboTP 9mo ago
Yes but in my experience this sometimes works great, other times you paint yourself in a corner and the sun total is that you still have to learn the thing, just the initial ram is less steep. For example I build my self a nice pipeline for
780.
by bonoboTP 9mo ago
I checked the link, it never says that the model's prediction get lower quality due to batching, just nondeterministic. I don't understand why people conflate these things. Also it's unlikely that they use smaller batch sizes

More ›