nl

46 ms·

781.
by nl 10mo ago
> The hikikomori[1] or NEETs ought to be a hotbed of creative works if your hypothesis is true. This seems a non-sequitur. People whose motivation is isolation are unlikely to try to generate anything for other people. But your general i
782.
by nl 10mo ago
Remember when 3D printers meant the death of factories? Everyone would just print what they wanted at home. I'm very bullish on LLMs building software, but this doesn't mean the death of software products anymore than 3D printer
783.
by nl 10mo ago
Generally they only load a bit of the project into the context at a time. Grep works really well for working out what.
784.
by nl 10mo ago
I wrote my own agent where everything happens over SSH. The shell is SSH, the read_file and write_file tool calls are over SSH Then I give it a disposable VM and let it go. There are lots of other solutions, but it's an interesting p
785.
by nl 10mo ago
> omniscient vibe... that HN human commenters usually don't au contraire! HN commenters are notoriously overconfident of their (our?) knowledge, especially in areas outside our expertise. > Do you think the root comment is _not_
786.
Tim Dettmers: A Personal Guide to Automating Your Own Work(timdettmers.com)
3 points by nl 10mo ago|0 comments
787.
by nl 10mo ago
OpenAI indemnifies customers against copyright claims: https://www.lexology.com/library/detail.aspx?g=671fdd7f-3cef...
788.
by nl 10mo ago
AMD Ryzen AI 300 These are great . The Ryzen AI series are the ones that allow memory to be shared between the GPU and CPU, so you can use almost all your system RAM to run local models. The AI 395+ MAX is available with up to 128Gb RAM
789.
by nl 10mo ago
Is there actually demand for this? I did one that writes to a Notion DB and is hosted for free on Cloudflare. I don't think I bothered to open source it though.
790.
by nl 10mo ago
Does Linus Torvalds count?
791.
by nl 10mo ago
I'm vibe coding a fantasy console with the feel of early 90s computing. 100% vibe coded WASM based. It runs Doom so far. I'm building that because I'm building a free-to-use vibe coding platform (ala Lovable) but the entire w
792.
by nl 10mo ago
This is wrong and I don't know why people think it is true. LLMs understand language Grammar files really well. A new language is easy for them (you can tell this by giving them a JSON schema and seeing how well they do) What they do
793.
by nl 10mo ago
> Lawyers have even started hiring junior lawyers back I assure you that this isn't anything like the level before. Lawyering has changed forever.
794.
by nl 10mo ago
Modern car brakes don't have asbestos. The difference in tyre wear is so marginal it's probably unmeasurable - less than the difference between running at the correct pressure and forgetting to check your tyre pressure. ICE vehicl
795.
by nl 10mo ago
No true Scotsman would ever do the kind of programming that 95% of the programming in the world is right? Anyway https://www.reddit.com/r/osdev/comments/1opsicd/just_how_far...
796.
by nl 10mo ago
I had it write a new stdlib implementation (wasm<->JS) for a custom web based fantasy console I'm writing. It did ok at that. We'll - Doom runs so ok enough for what I wanted anyway. No, it's not a copy of other WASM st
797.
by nl 10mo ago
> are 99% of us building a pacemaker? no. but that doesn’t mean we can automatically make the leap to assuming a set of tools known for being non-deterministic are good enough for our use case. This seems irrelevant? Either way hopefully
798.
by nl 10mo ago
> Assuming the LLM in use has, or is reasonably expected to have, model evolution, documents generated by same will diverge unpredictably given a constant prompt. So what? You tell it once. It writes code. You test that code, not the pro
799.
by nl 10mo ago
> Using English, with all its inherent ambiguity, to attempt to communicate with an alien (charitably) mind very much does /not/ make this task any easier if the thing you need to accomplish is of any complexity at all. This ju
800.
by nl 10mo ago
I love building things with computers. I'm not particularly interested in the coding. I've coded professionally for 30 years (ergh!). I'm ok at it. But I love building things with AI. I haven't had this much fun since
801.
by nl 10mo ago
> Compilers are deterministic, making their generated assembly code verifiable People keep saying this like it is an absolute fact, whereas in reality it is a scale. Compilers are more deterministic than LLMs in general, but no they are
802.
by nl 10mo ago
What are the exact political views the Cloudflare is pushing here? That it is unreasonable for Italian soccer rights owners to try to use Cloudflare to enforce their broadcast restrictions with 30 minutes notice? That it is unreasonable not
803.
by nl 10mo ago
> I suspect it’s closer than you think! It's not. I've done this (although not with all these tools). For a reasonable sized project it's easy to tell the difference in quality between say Grok-4.1-Fast (30 on AA Coding In
804.
by nl 10mo ago
This is SO wrong. I actually wrote my own simple agent (with some twists) in part so I could compare models. Opus 4.5 is in a completely different league to Sonnet 4.5, and 3.7 isn't even on the same planet. I happily use my agent wi
805.
by nl 10mo ago
This is exactly my point.
806.
by nl 10mo ago
I don't think anyone has ever used it as training. But yes labs still do seem to target it as goal (which is a different thing).
807.
by nl 10mo ago
Say I want to auto-bookmark a bunch of tabs and need a summary of each one. Using the title is a mechanical solution, but a nice prompt and a small model can summarize the title and contents into something much more useful.
808.
by nl 10mo ago
Yes, I thought that too! But qwen3:0.6b (and to some extent gemma 1b) has made me reevaluate. They still aren't useful like large LLMs, but for things like summarization, and other tasks where you can give them structure but want the s
809.
by nl 10mo ago
I've been super impressed by qwen3:0.6b (yes, 0.6B) running in Ollama. If you have very specific, constrained tasks it can do quite a lot. It's not perfect though. https://tools.nicklothian.com/llm_comparator.html
810.
by nl 10mo ago
I wrote https://tools.nicklothian.com/llm_comparator.html so you can compare different models. OpenRouter gives you $10 credit when you sign up - stick your API key in and compare as many models as you want. It's all b

More ›