benatkin

11 ms·

151.
by benatkin 8mo ago
I'd like a reference for it being rug pulling . What happened with OpenCode certainly wasn't rug pulling , unless Anthropic asked them to support using a Claude subscription with it.
152.
by benatkin 8mo ago
But they're separate highly maintained projects, and there will always be tradeoffs. It's like saying that Ubuntu is better than Debian, or that Fedora is better than RockyLinux.
153.
by benatkin 8mo ago
Could say the same thing about people working on neovim
154.
by benatkin 8mo ago
DeepMind wrote the paper, and while Google's API arrived later than OpenAI's it isn't as late as some people think. The PaLM API was released before the Gemini brand was launched. Microsoft funded OpenAI and popularized ear
155.
by benatkin 8mo ago
What's the purl (Package URL) equivalent of surl:mystery, for stuff like Claude Code, which now only supports running a script to install? It does have a pretty easy to read install script, but the docs don't suggest reading it be
156.
by benatkin 8mo ago
Don't worry, author, I don't think you're a luddite. You make that quite clear with this: > I can't imaging writing code by myself again After that, you say that you need to know the intention for "content".
157.
by benatkin 8mo ago
Each article like this one is an opportunity to assess whether it's mainly written by an AI or not. After reading part of this one I mostly think not (except for the obvious AI generated image), but it would be amusing if it were. &quo
158.
by benatkin 8mo ago
Basketball is a general purpose sport. The Claude of it can win. Some other sports such as gymnastics would need something more like the AlphaZero of it to win.
159.
by benatkin 8mo ago
That was covered just fine IMO. The reaction seems to be "so what?" I think that's a valid reaction. It's a long article to state something obvious, that the important thing about being on your way to greatness is having
160.
by benatkin 9mo ago
Needs to be a new HTTP status code to go along with 418. Never mind that it doesn't start with the right number. Also I'm sure glad that scammer didn't manage to buy that cannon!
161.
by benatkin 9mo ago
Try it again. This time do something different with CLAUDE.md. By the way it's happy to edit its own CLAUDE.md files (don't have an agent edit another agent's CLAUDE.md files though [0]) 0: https://news.ycombinator
162.
by benatkin 9mo ago
> betting actual money on those claims Insurance companies can let marketing influence rates to some degree, with programs that tend to be tacked on after the initial rate is set. This self driving car program sounds an awful lot like sa
163.
by benatkin 9mo ago
The readme exaggerates the threat of agents shelling out and glosses over a serious drawback of itself. On the shelling out side, it says "One prompt injection and you're done." Well, you can run a lot of these agents in a co
164.
Software Survival 3.0(steve-yegge.medium.com)
6 points by benatkin 9mo ago|0 comments
165.
by benatkin 9mo ago
> AI SRE in 2026 is what DevOps Anomaly Detection was in 2015 — bold claims backed by huge marketing budgets, but lacking independent verification. There are stories of SaaS vendors abruptly killing the observability stack. Our results m
166.
by benatkin 9mo ago
Good time to surface the limitations of a Content Security Policy: https://github.com/w3c/webappsec-csp/issues/92 Also the double iframe technique is important for preventing exfiltration through navigation,
167.
by benatkin 9mo ago
Yes, and with GPU, it's Whisper, which has been mentioned elsewhere in this article's comments. I mean that handy.computer provides the other option as a fallback for those who can't or don't want to use the GPU.
168.
by benatkin 9mo ago
To save a click, it's just a fancy front end for Whisper plus a weaker CPU-only model. It has a demo video that seems impressive, but the speech is careful to sound casual while having no meaningful flaws that would cause it to mess up
169.
by benatkin 9mo ago
I'm going to propose a law for these AI orchestration systems based on Greenspun's Tenth Law: > Any sufficiently complicated AI orchestration system contains an ad hoc, informally-specified, bug-ridden, slow implementation of h
170.
by benatkin 9mo ago
To me, this sounds more monolithic than containers. I think I'd like something less monolithic. However, those who like monorepos might be more quick to develop an interest in this. I could of course use containers within the Micro
171.
Cursor may be switching from Solid to React(cursor.com)
3 points by benatkin 9mo ago|1 comments
172.
by benatkin 9mo ago
https://x.com/brenelz/status/2011598823244890409
173.
by benatkin 9mo ago
It also shows that it isn't perfectly organized, that it isn't an ideal model for knowledge aggregation. If it's ideal for it to be globally consistent, then it doesn't have that. If it's ideal for it to be adapted
174.
by benatkin 9mo ago
Found yours while searching for "llm"
175.
by benatkin 9mo ago
I propose the name tokables for the compressed data produced by this. A play on tokens and how wild it is.
176.
by benatkin 9mo ago
Peoples' writing is influenced by what they read, so such a strong objection to someone suggesting that an LLM might have been involved in the text of a blog post won't fly with me.
177.
by benatkin 9mo ago
I wondered if maybe it was about Vienna
178.
by benatkin 10mo ago
I wanted to say I that I think it's overrated in terms of its position on HN, but rather than criticize side issues of it, which often point to something being a weak article in general, I probably should have just said exactly what I
179.
by benatkin 10mo ago
There are also no citations, and it has this phrase "This website is not licensed for ML/LLM training or content creation." Yeah right, that's like the privacy notice posts people make to facebook from time to time that
180.
by benatkin 10mo ago
Turing test type stuff. Having agents try to participate in the overall social community, by sending a cold email. The reason I think it's good is because with how AI is increasingly used a lot of different sorts of research is needed.

More ›