These posts make me feel like I’m the worst llm prompter in existence.

I’m using a mix of Gemini, grok, and gpt to translate some matlab into c++. It is kinda okay at its job but not great? I am rapidly reading Accelerated C++ to get to the point where I can throw the llm out the window. If it was python or Julia I wouldn’t be using an LLM at all bc I know those languages. AI is barely better than me at C++ because I’m halfway through my first ever book on it. What LLMs are these people using?

The code I’m translating isn’t even that complex - it runs analysis on ecg/ppg data to implement this one dude’s new diagnosis algorithm. The hard part was coming up with the algorithm, the code is simple. And the shit the LLM pours out works kinda okay but not really? I have to do hours of fix work on its output. I’m doing all the hard design work myself.

I fucking WISH I could only work on biotech and research and send the code to an LLM. But I can’t because they suck so I gotta learn how computer memory works so my C++ doesn’t eat up all my pc’s memory. What magical LLMs are yall using??? Please send them my way! I want a free llm therapist and a programmer! What world do you live in?? Let me in!

A lot of people are using Claude Code which many consider to be a noticeably better for coding than the other models.

I think also they tend to be generating non-C++ code where there are more guardrails and less footguns for LLMs to run into. Eg they're generating Javascript or Python or Rust where type systems and garbage collection eliminates entire classes of mistakes that LLMs can run into. I know you said you don't use it for Python because you know the language but even experienced Python devs still see value in LLM-generating Python code.

That’s funny bc I linked my post to a server I’m on and I also was told to use an agent.

My worry about an agent is I’m trying to translate the math with full fidelity and an agent might take liberties with the math rather than full accuracy. I’m already having issues with 0 to 1 indexing screwing up some of the algorithm.

But I will try an agent - can’t hurt to try

I'm firing you for being unable to adequately commune with the machine spirit.

(But for real, a good test suite seems like a great place to start before letting an LLM run wild... or alternatively just do what you're doing. We definitely respect textbook-readers more than prompters!)

I’m a contract hire that you paid upfront! You can’t fire me!

Also let this be a lesson to internet folks to be careful what you post if your boss shitposts on the orange yelling site

The type of content I come to HackerNews for

See I like this but also the majority of software jobs these days are a net negative to humanity

Maybe I’m just a silly lab engineer who doesn’t know how big companies work (i did work at a startup for a time but that’s even more anti meeting than a lab) but I feel like maybe if you have so many back to back meetings that you need to plan around them you should have fewer meetings?

Also why is there a hard end time rather than a maximum time? What is even going on? How are you getting work done?

At large companies, there's just a lot of coordination needed. Which means a lot of meetings typically. Probably more than there is needed but one of the points of a large company is having a critical mass of people pulling in roughly the same direction(s).

>also why is there a hard end time rather than a maximum time?

Depends on company, culture, and individuals. If you get through the purpose of the meeting, the meeting organizer may well say "Let's give everyone 20 minutes (or whatever) back." I do think that it's a worthwhile goal to default to <1 hour meetings.

In any case, agendas and thinking about how the meeting time should actually be spent is worthwhile.

At least at my multinational company, we have teams iny department that are 12 hours ahead of us. That makes the window of overlap time we have small, so we often have blocks of meetings during that overlap time that compete for time slots for staff that make higher level decisions.

I really love Zed bc it’s fast and doesn’t try to run the code for you. I really dislike the “play” feature of IDEs. I learned Java in eclipse and for the linear time don’t know how to run it from the terminal. It has all the editing and highlighting features of a full ide but makes you run it from the terminal. And if you have Zellij or another muxer set up, it runs it in the Zed terminal automatically.

I like metaphors. I respect a good “run with a metaphor” post. But I feel like if you have to explain the metaphor this much during the post it’s not worth it

True. This particular lily was less gilded than battered and deep fried.

"This post could have been a tweet", and so on.

It's a lovely metaphor for the soulful business of doing something really well, even if there is no 'market' for that kind of quality

Art is often its own reward.

I feel like my learning curve for AI is:

1) Learn backprop, etc, basic math

2) Learn more advanced things, CNNs, LMM, NMF, PCA, etc

3) Publish a paper or poster

4) Forget basics

5) Relearn that backprop is a thing

repeat.

Some day I need to get my education together.

Today I learned that there is such a thing as an array language! This is so neat! Need to learn all of them.

Start here: https://xpqz.github.io/learnapl

or here:

https://xpqz.github.io/kbook

I work in a lab which researches regenerative medicine and people often talk about how we are saving lives.

But at the end of the day, it’s not what we are doing that is pushing the needle, its projects like this make existing healthcare accessible for everybody. I want to get a PhD in something related to this at some point. You guys are so insanely cool.

Maybe I live in a bubble but these are not problems I’ve ever heard expressed until now. Bluesky seems perfectly fine for those who use it?

Everyone I actually enjoyed following is on there. Someone once posted that the best part of Bluesky is that you scroll until you reach the end of the new posts from users you follow, then once you reach the end of new content you stop and go do something else for a while. Bluesky not attempting to be a infinite dopamine is a feature in the best interest of the users

Exactly. The best replacement for Twitter/Reddit/etc. isn't any single option, it's your favorite of the new options plus a mindset of being less online, not needing a replacement for how important those used to be.

I’ve been using RSS feed for nearly 20 years now. Why we ever strayed to some endless algorithm is beyond me.

I have a list of chronologically sorted articles from sources I trust. That’s it and that’s all.

I read them until they are read. Then I close the app and do other stuff until tomorrow.

Is it any wonder Google killed off Reader around the same time it tried to launch Google+? Managing our own feeds was never going to peak capitalism.

Maybe it’s fine if you treat Bluesky as a way to follow just a few trusted people. But that could also be an echo chamber. I was surprised, however, by how many vile comments I saw regarding Charlie Kirk’s death. To me as an outsider to its culture, it looks just like Reddit. So “sectarian” feels right.

I use blue sky exclusively over Twitter and in my experience the condolences about kirks death out numbered the vile posts at least 10 to 1. In general the worst I saw were posts lamenting his death by violence while also noting his bigotry and love for guns.

I didn't care much for it as it seemed that anything other than a left-leaning anti-american view was complexly absent and unwelcome. I kind of miss the google circles thing... it was much easier to separate technical or career vs personal vs political.

> Why is Bluesky an echo chamber but X, Rumble, Gab, Gettr are not? I bet there's huge vile "damn leftists" commentary there, even though we don't know who shot Kirk or why. Why isn't that an echo chamber?

I have not even heard of some of those. So I am not saying that they aren’t echo chambers either. But given the reputation for civil discourse that blue sky has, I was surprised to see a lot of comments that were saying disgusting things in thinly coded ways.

Why is your experience better than mine? I didn't see those things.

I didn’t say it was?

It seems logical that people who like it keep using it, while people who don't like it stop. I've seen several people right here on HN say they stopped using it for various reasons, so I don't suppose he's making it up.

Interesting that there roughly 700K daily posters but only 390K followers. I have no idea what other social media numbers are but having ~2x the posters as followers doesn't seem like a sign of health to me.

You are misinterpreting the statistics.

Having more people talk than follow is good thing and probably consistent across social media.

That's just saying the average person posts for 2 days and follows for 1. Which seems very typical if not heavy on the following.

Pointing at more people speaking than sheeping and calling that an ominous statistic seems off, don't you think?

It's over twice as popular as it was a year ago.

There was an election combined with it being new - now it's trending to be the next Clubhouse. Remember Clubhouse?

Bluesky looks to go up then fall back to about half of it's peak. It has the look of a consistently used platform that slowly rolls off users but maintains a solid base. Clubhouse went to the moon then died over a 6 month period. Bluesky would need to drop another 2/3 of it's active users 6 months ago if you wanted your statement to match your data. Bluesky isn't growing exponentially, and is not falling catastrophically. Not a good comparison.

I see a slight decline of around 15% after the initial surge followed by relatively stable activity. It must be easy to manipulate someone who sees stats through such warped glasses. They are playing you like a fiddle. Or maybe you're just trying to bias other folks.

No, it's down posts year over year, which I wouldn't have believed was possible if it hadn't happened. It's got half as much post traffic as its peak(s), which were on exactly the days that Trump was elected and inaugurated.

It's basically what Truth Social would be if it didn't even have Trump.

-----

edit: it's not the fault of the technology, it's the fault of the awful company. They represent people strongly aligned with the Democratic Party, but with no power in the Democratic Party, and their philosophy has been stated over and over again: if you don't agree with whatever we believe today, we don't even want your support, or want you here, or want you to be employed. A lot of conservatives would add "they also don't want us to be alive."

The worst part is that they're all upper-middle class, and when as the Gini coefficient goes up with the right-wingers they're locking into power by being so repulsive, they'll just get wealthier and wealthier and more self-righteous.

I've been suspecting for years that there's a lot of botted support for the dumbest most out of touch liberals that is paid for by conservatives. I don't meet people like this in real life, and I am very left-wing. The liberals I meet are generally humble and thoughtful (if in love with their television sets.)

I think bsky is a good platform in part because people like you just bounce off lol