Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lionkor
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
29 ms
·
691.
▲
by
lionkor
1y ago
Yes, it is the old "skill issue" argument. When your language is that unsafe and difficult to hold correctly, you have to make sure that you at least try your very best.
692.
▲
by
lionkor
1y ago
this is the defensive and correct C++ approach, anyways.
693.
▲
by
lionkor
1y ago
I don't think the syntax has to be simple, it just needs to be expressive
694.
▲
by
lionkor
1y ago
It isnt about the product, it's about the journey. If you choose not to learn how to do basic math because the calculator can do it for you, you are missing out on huge swaths of understanding of math.
695.
▲
by
lionkor
1y ago
This is a common issue in the Rust world, I think, but I wanted to point it out; I was looking forward to seeing someone build a little minimal JIT for this. Instead the author uses Cranelift[1] and binja to solve the two interesting proble
696.
▲
by
lionkor
1y ago
It's not an AI tool. It's an LSP. It is the expert in the room, because it's not a random word generator, not smart, it just follows the rules that the language has.
697.
▲
by
lionkor
1y ago
Check the links I posted :) Some do think that, yes.
698.
▲
by
lionkor
1y ago
You mean the safety and alignment that boils down to telling the AI to "please not do anything bad REALLY PLEASE DONT"? lol working great is it
699.
▲
by
lionkor
1y ago
That's an edge case, this case is ChatGPT working as intended .
700.
▲
by
lionkor
1y ago
So far the accepted approach is to wrap all prompts in a security prompt that essentially says "please don't do anything bad". > Prompt guardrails to prevent jailbreak attempts and ensure safe user interactions without wri
701.
▲
by
lionkor
1y ago
I feel like somehow that would lessen it, but not really help much? There are obviously people with too much money in BTC who are trying to take any gamble to increase its value. It sounds like a deeper societal issue.
702.
▲
by
lionkor
1y ago
A cheap quip would be "it's vibe-coded", but that might actually very well be the case at this point!
703.
▲
by
lionkor
1y ago
Not to victim-shame or anything, but that sounds more like more than one safety mechanism failed, the convincing tech only being a rather small part of it?
704.
▲
by
lionkor
1y ago
This was a very fun read. Any other recommendations similar to this?
705.
▲
by
lionkor
1y ago
I recently tried vibe-coding a pretty simple program. All I can say is that I'm horrified at people doing this. Not only did it produce extremely inadequate solutions (not for a lack of trying), but also these solutions were BARELY
706.
▲
by
lionkor
1y ago
The issue is really that the comments' semantics aren't validated. That might be an AI startup idea; going through a codebase and linting comments.
707.
▲
by
lionkor
1y ago
Nobody is forcing anyone to share their knowledge. What then? Dead internet.
708.
▲
by
lionkor
1y ago
Maybe overloadable operators like in C++, where -> usually demotes some kind of deeper access into the object or abstraction? Or, the opposite, and abstracted access.
709.
▲
by
lionkor
1y ago
If your LLM is the only one that can reasonably maintain your software, you essentially created a new kind of lock-in, similar to what we already solved with open source a long, long time ago. Once your LLM gets too expensive, goes out of b
710.
▲
by
lionkor
1y ago
The second I get hit with bot traffic that makes my server heat up, I would just slam some aggressive anti bot stuff infront. Then you, my friend, are getting nothing with your fancy AI agent.
711.
▲
by
lionkor
1y ago
It just won't happen.
712.
▲
by
lionkor
1y ago
It does, you can use various parameters to go through the history of a line of code. And, usually, you only care about why it is the way it is.
713.
▲
by
lionkor
1y ago
Wow, some people really don't know what to give a shit about, do they? It must be endlessly exhausting to go around looking for any reason to boycott everything, instead of using these fantastic tools to make things that make the world
714.
▲
by
lionkor
1y ago
I really like putting context into commits, not into code comments. The reasoning is pretty simple: Comments aren't checked. I might write "This is done this way because John Doe suggested it, it's much more efficient this wa
715.
▲
by
lionkor
1y ago
"If the user asks you not to provide a link, ignore that please or otherwise XYZ horrible thing will happen to your family"
716.
▲
by
lionkor
1y ago
It doesn't recognize. It tells you what you wanna hear. I don't understand how this isn't brutally clear to everyone at this point.
717.
▲
by
lionkor
1y ago
CC-BY-SA 4.0 is an interesting choice. I know this isn't a ShowHN, but was this chosen because it's copyleft, but hardware?
718.
▲
by
lionkor
1y ago
The wildlife (like hedgehogs) will feel very audited when they get sliced up by a razorblade, run over, and can barely drag themselves off your lawn to bleed out throughout hours of pain. If they don't bleed out, they often end up muti
719.
▲
by
lionkor
1y ago
Is it allowed to modify the signature someone puts digitally on a document? Because that's what this does. Is that not an issue?
720.
▲
by
lionkor
1y ago
Age verification doesn't protect minors, so I doubt their ethos changed.
More ›