Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
solaire_oa
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
31.
▲
by
solaire_oa
7mo ago
Thanks! Nice, you found the alltrust.json file ha. Yes, a bg job running on an rpi leverages HN APIs and builds the alltrust file by the minute, for all "active" accounts. Technically fetching that data is all you'd need to m
32.
▲
by
solaire_oa
7mo ago
Oh wow, You're a human with a high submission rate! I assumed accounts like these were bots. I've seen a large number of accounts like yours in `/newest`, indeed, it's the reason I made the default demo penalize high sub
33.
▲
by
solaire_oa
7mo ago
Author here. This is a good callout, there are a few reasons why it's a plugin and not open source (yet). First is that I didn't want to make a plugin in the first place, I wanted to make a bookmarklet, but HN's CSP policy wa
34.
▲
by
solaire_oa
7mo ago
Author here. I tried a bookmarklet, that was my preference, but the security headers in HN were too strict. As far as I could tell anyway.
35.
▲
by
solaire_oa
7mo ago
This points out that agentic security flaws are worse than "systemic", they're the feature. Agents are literal backdoors. It's so bizarre to be discussing minor security concerns of backdoors, like trying to block env va
36.
▲
Trust signals as sparklines for Hacker News
(hn-trustspark.com)
82 points
by
solaire_oa
7mo ago
|
46 comments
37.
▲
Pascail's Wager
(r5d.me)
3 points
by
solaire_oa
7mo ago
|
0 comments
38.
▲
by
solaire_oa
7mo ago
Related https://github.com/anomalyco/opencode/issues/10416
39.
▲
by
solaire_oa
7mo ago
Fair point. I figure that the translator itself is a small LLM. It's humorous though, hence the upvotes- dunking on LinkedIn easy and cathartic.
40.
▲
by
solaire_oa
7mo ago
Small team, backend. NDAs prevent widespread LLM use, but some of our engineers, junior and senior both, feel pretty confident in using Claude for "isolated" development, like generic packages or libraries that are plausibly unrel
41.
▲
by
solaire_oa
7mo ago
Similarly, managers at my workplace occasionally use LLMs to generate jira tickets (with nonsense implementation details), which has led junior engineers astray, leaving senior engineers to deal with the fallout.
42.
▲
by
solaire_oa
7mo ago
You've described a conference as well, for what that's worth.
43.
▲
by
solaire_oa
7mo ago
I'm late to the party, but this seems like the answer to the slop-pocalyspe!
44.
▲
by
solaire_oa
7mo ago
I am having a blast with this: https://github.com/samuelclay/hackersmacker I dropped into /new so that I can "foe" all the barf.
45.
▲
by
solaire_oa
7mo ago
I created a Firefox plugin which takes HN commenters'/submiters' account create date and sort/scales the order/points based on its creation since 2009 (older accounts get more weight). Optionally, the plugin just pu
46.
▲
by
solaire_oa
8mo ago
If it helps, I would consider learning about and deploying local LLMs (llama.cpp etc), if you have the hardware (and it doesn't take all that much hardware at all). They're nowhere near as fast as LLM providers, but they come fair
47.
▲
by
solaire_oa
8mo ago
This is funny, but I do feel like I just got bait-and-switched, where I was hoping for a non-joke protocol.
48.
▲
by
solaire_oa
8mo ago
We should definitely feel trepidation at the prospects of any LLM guided browser, in addition to WebMCP (e.g. Claude for Chrome enters the same opaque LLM-controlled/deferred decision process, OpenClaw etc). Just one example: Prompting
49.
▲
by
solaire_oa
8mo ago
To clarify, you mean that we're entering a post-HTML world, correct? As in, why spend effort on the aesthetics if a human will never see it, correct? Because that is also my worry; a post-HTML and perhaps even a POST-API world....
50.
▲
Vibeschism
(r5d.me)
2 points
by
solaire_oa
8mo ago
|
0 comments
51.
▲
by
solaire_oa
8mo ago
Additionally, the boredom atrophies any future collaboration. Why make a helpful library and share it when you can brute force the thing that it helps? Why make a library when the bots will just slurp it up as delicious corpus and barf it b
52.
▲
Vibeschism
(r5d.me)
3 points
by
solaire_oa
8mo ago
|
0 comments
53.
▲
by
solaire_oa
8mo ago
Music? Can we even call it that? Sounds created by LLMs are one big If By Whiskey Speech. If by music you mean sounds that fill the silence, pad videos with unearned pathos, fill elevators to prevent strangers from feeling awkward or talkin
54.
▲
by
solaire_oa
8mo ago
This is an excellent step in the direction of a web-of-trust that the present moment demands, facing an increasingly mistrustful web in the face of LLMs. Major congratulations to the creator, you're doing god's work. And even if t
55.
▲
by
solaire_oa
9mo ago
It would be amusing if somebody parodied this article titled "AI is killing speculative articles about AI".
56.
▲
by
solaire_oa
9mo ago
Oh, having a historian's translation of the essay is legitimately helpful for understanding how the the points relate in the current climate, thank you. And I love acoup.blog, separately (tangentially related, I was legitimately crestf
57.
▲
by
solaire_oa
9mo ago
To be honest, I found this essay a bit meandering and I lost interest in reading the whole thing (coming off the heels of re-reading Civil Disobedience, which by comparison is immediately forthright and to the point). Skipping ahead to the
58.
▲
by
solaire_oa
9mo ago
You're being downvoted, I think unfairly, because this is a completely valid rebuttal. Without going into a diatribe about how governments necessitate coercion and violence, enforcing such a "law" would indeed be counterprodu
59.
▲
by
solaire_oa
9mo ago
As much as I agree with this sentiment and think it's poignant, civil conflict in the modern era would be unthinkably terrifying, so I wouldn't take this position lightly. Or at the very least I wouldn't compare a modern conf
60.
▲
by
solaire_oa
9mo ago
Which isn't to say that "making commerce cheaper in Europe" isn't a worthwhile goal, it is- but at face value this attempt looks comically inept. "We wrote an email to a politician! Progress due any moment! Like and
More ›