solaire_oa

10 ms·

61.
by solaire_oa 9mo ago
I've been thinking this exact thing! But it's too abstract a thought for me to try creating anything yet. A curation network, one which uses SSL-style chain-of-trust (and RSS-style feeds maybe?) seems like it could be a solution,
62.
by solaire_oa 9mo ago
I think this is a fair take (despite the characteristic HN negativity/contrarianism), and succinctly summarizes a point that I was finding hard to articulate while reading the article. My similar (verbose) take is that seniors will oft
63.
by solaire_oa 10mo ago
I wasn't aware of Droidian, this looks great, thanks!
64.
by solaire_oa 10mo ago
I can confirm, this is sensible advice. I backed the FxTec Pro 1x during a depressive period in depths of COVID. It took ~years of hamstringing for them to deliver, but they did eventually deliver the phone. Even aged as the phone is, it&#x
65.
by solaire_oa 10mo ago
Most people in this thread are quibbling about the exact degree of utility LLMs provide, which a tedious argument. What's more interesting to me is, per the article, the concern regarding everyone who is leaning into LLMs without reali
66.
by solaire_oa 10mo ago
I would summarize Americans (and perhaps most English speaking countries) as perceiving this mindset to be callous, ineffective, and a dereliction of autonomy. But I'm interested in how Germans perceive Americans in reverse? If shop st
67.
by solaire_oa 10mo ago
That's great! I'm not accusing you or anything, or myself for that matter, I'm only lamenting that open source will likely decay due to LLMs. Or at the very least, open source will no longer be seen as the height of virtue in
68.
by solaire_oa 10mo ago
First, I'll second that I've applied agentic LLMs to an open source project to fix bugs and forcibly coerce it to act in ways that the maintainer may or may not approve of. It has been remarkablely effective, so long as I'm w
69.
by solaire_oa 11mo ago
You say that you can't really agree [about the resume being poorly formatted from having too much text?], but then you agree that there's too much text (if all you need to see is the 1 item of "Google", then you're
70.
by solaire_oa 1y ago
The only compelling reason for me to try Gemini CLI opposed to Claude Code is that Gemini CLI supposedly supports ollama, which could potentially open up avenues for locally hosted agentic coding (lower quality, but private). When I attempt
71.
by solaire_oa 1y ago
I can second the recommendation for FreshRSS- it's incredibly utilitarian and I have no complaints. I've been using it for about 4 years. It's how I came to this thread! Couldn't agree more about the "final version&
72.
by solaire_oa 1y ago
As disgusting as this decision is, hopefully it de-emphasizes the idea of mobile apps in general, which (besides CPU heavy apps like video games) are often web-apps with extra steps (React/Ionic/WebView/etc).
73.
by solaire_oa 1y ago
This happened for me as well. It adds insult to injury that lying awake, unable to sleep properly for months straight, doomscrolling (or just being online too much) further saps your mental health in an already drained and depressed state.
74.
by solaire_oa 1y ago
> The sky above the port was the color of television, tuned to a dead channel. > static-filled “dead channels.” I don't think Gibson was referring to static (which is bland grey and not cyberpunk at all). I think he was referring
75.
by solaire_oa 1y ago
Revisiting code review in terms of how it functioned in 2020 seems antiquated. Security and quality are a concern now that there's a flood of LLM barf that inexperienced engineers are liable to submit for code review. Code review has s
76.
by solaire_oa 1y ago
Ok. Tried it, I'm not super impressed. Client: Ollama (phi4) - 90164ms. StopReason: stop. Tokens(in/out): 365/396 ---PROMPT--- user: Extract from this content: Grave Digger: Ingredients - 1 1&
77.
by solaire_oa 1y ago
Interesting! I was using function calling in OpenAI and JSON mode in Ollama with zod. I may revisit the project with SAP.
78.
by solaire_oa 1y ago
I also think that structured outputs are criminally underused, but it isn't perfect... and per your example, it might not even be good, because I've done something similar. I was trying to make a decent cocktail recipe database, a
79.
by solaire_oa 1y ago
Top-tier professional programmer quality is exceedingly, impractically optimistic, for a few reasons. 1. There's a low probability of that in the first place. 2. You need to be a top-tier professional programmer to recognize that type
80.
by solaire_oa 1y ago
Making 4 PRs for a well-known solution sounds insane, yes, but to be the devil's advocate, you could plausibly be working with an ambiguous task: "Create 4 PRs with 4 different dependency libraries, so that I can compare their imp
81.
by solaire_oa 1y ago
IIUC one reason is that prompts and other data sent to 3rd party LLM hosts have the chance to be funneled to 4th party RLHF platforms, e.g. Sagemaker, Mechanical Turks, etc. So a random gig worker could be reading a .env file the intern upl
82.
by solaire_oa 1y ago
Not to make you more downtrodden, but it's not like AI would have any trouble at all producing a small page. If you mean that there needs to be signals in place that an article was thought about and physically typed up by human fingers
83.
by solaire_oa 1y ago
Person A as can commit atrocious code all day, that's fine, but they still need to proofread their MR/PR and fix the outstanding issues. The only way to see outstanding issues is by reviewing the MR/PR. Good writers proofread
84.
by solaire_oa 1y ago
Writing/reading code and reviewing code are distinct and separate activities. It's completely common to contribute code which is not production ready. If you need an example, it's easy to add a debugging/logging statemen
85.
by solaire_oa 1y ago
The submitter should also have thoroughly reviewed their own MR/PR. Even before LLMs, coders not having reviewed their own code would be completely discourteous and disrespectful to the reviewer. It's an embarrassing faux pas that
86.
by solaire_oa 1y ago
Agreed, the messiness of the original text has character and humanity that is stripped from the summarized text. The first text is an original thought, exchanged in a dialogue, imperfectly. Elsewhere in this comment section, it's disc
87.
by solaire_oa 2y ago
It's not directly related to the article, but the music video for "Moving On" by James is singularly the most moving video I've seen about coming to terms with death and birth. https://www.youtube.com/wat
88.
by solaire_oa 2y ago
Better how? And by what measurement? What are you gaining from info overload opposed to ignorance? You say you fundamentally believe in one over the other, but you haven't made the case for info overload, and have even made some points
89.
by solaire_oa 2y ago
Tangentially related, using profanity in text is a pretty fucking strong way to signal that the text wasn written by an actual human.
90.
by solaire_oa 2y ago
I've been using FreshRSS for about 3 years now, with pretty much the same sources of feeds as the author. The main pro is that the signal to noise ratio is incalculably better than social media. It's trivial to obtain updates you&

More ›