toraway

7 ms·

31.
by toraway 4mo ago
Possibly a bug, but the change in usage quotas on my AI Pro plan going from Gemini CLI to Antigravity CLI was a massive drop. I was kicked off after around 30 minutes using the smallest model available (3.5 Flash). If they offered 3 Flash (
32.
by toraway 4mo ago
Thank you for sharing, that was one of the most insightful long form pieces I've read in a long time! And the writing was enjoyable to read even as a math layperson. I was going to say you should submit it but I saw you did a few days
33.
by toraway 4mo ago
Jellyfish has improved a bunch in the last few years, the front end is a lot more polished. I finally moved on from Plex a year ago after some kind of upsell nag on a basic feature after already paying for a plan. Although realtime detectio
34.
by toraway 4mo ago
That's a lot of hyperbole, there's no cause/effect relationship I can think of here that could realistically produce your slippery slope. Google or anyone else could start adding those unique tracking watermarks you're c
35.
by toraway 4mo ago
FWIW there are a few people in the issues saying that the tool is giving false negatives and the output image gets flagged by the actual Gemini API as having SynthID. Most recently 3 weeks ago without a response.
36.
by toraway 4mo ago
That would be Flash Lite now, and I'm also interested in the cheaper end of things so kinda disappointed they didn't release 3.5 Flash Lite at the same time...
37.
by toraway 4mo ago
> The result is that the illegal market dwarfs the legal market. The legal suppliers simply can't compete with efficient and untaxed illegal or grey market sellers. Source for this? I mean consumer buying habits not illegal
38.
by toraway 4mo ago
Okay, but none of that is actually responsive to what the article is discussing, which is competitive CTFs. There's not a single criticism of using AI for actual security research in anything they wrote and they mention being a heavy u
39.
by toraway 4mo ago
Just parachuting in to reflexively throw the "Luddite" label at someone lamenting the decline of a niche community they've enjoyed participating in and contributing to is certainly ... a choice. Within the framework of your a
40.
by toraway 4mo ago
> Vietnam - unnecessary war, but we won the peace I’m struggling to understand what this spin is even supposed to mean? > Afghanistan - we tried our best and made some mistakes along the way. Eventually got Bin Laden thoug
41.
by toraway 4mo ago
From the excerpt it sounds like the author is just describing one specific archetype from within a list of others included in the book and doesn't make any claims about it being a uniquely common type within every org, or the most comm
42.
by toraway 4mo ago
The disclosure about being a honeypot is in the CONTRIBUTING.md: Warning Heads up: This is a research project — bounties listed here are symbolic and part of an academic study on open-source contribution patterns. PRs are reviewed
43.
by toraway 4mo ago
I recently left my mom a voicemail saying happy Mother’s Day with normal human boilerplate of sorry I missed you, feel free to give me a call back tonight or we can talk tomorrow, either is fine by me whatever works best for you, hope we ca
44.
by toraway 5mo ago
Also, as someone who has developed an ever growing suite of bespoke tools for my personal workflows using Codex/Gemini CLI over the last year, something I don’t see mentioned as often is the “mental overhead” of self-designed apps. Eve
45.
by toraway 5mo ago
Battery management tends to be best-in-class on Chromebooks, it's far from certain that you'll get anything nearly as good after installing 3rd party Linux on it. That's my #1 reason for not even considering it (despite havin
46.
by toraway 5mo ago
Not exactly "burying the lede" since Daniel already posted an update about it months ago [1] with extensive discussion in numerous of articles [2] including on this site [3]. [1] https://lists.haxx.se/pipermail&#x2
47.
by toraway 5mo ago
Also, the people at Mozilla who helped achieve a highly visible collaboration with the hottest AI company in the zeitgeist that included a lot of expensive data center time to harden their flagship product are definitely going to be happy&#
48.
by toraway 5mo ago
I've seen this suggested a few times in this thread but it seems like it's exactly backwards. Wouldn't that make it a better to distinguish whether Mythos is uniquely super powerful vs an incremental improvement from Opus etc
49.
by toraway 5mo ago
You could do that, but now you're like 90% of the way to maintaining a monotonically increasing number you that could just use as a unique ID instead without any randomness required (and without the additional 128 bits for collision pr
50.
by toraway 5mo ago
Shouldn't your test follow the pattern of how rng() is actually being used in the uuid.ts code internally? Your test is more-or-less contrived to fail given the tradeoff to avoid repeated memory allocations but that doesn't say mu
51.
by toraway 5mo ago
The problem with AI written articles is still feeling uncertain whether there's actually any utility after reading 2000 words as you realize that it's been 90% filler so far but think maybe it will lead somewhere soon? But it do
52.
by toraway 5mo ago
Exactly. This is the "Declare fentanyl a WMD" of solutions to ransomware. Sounds kinda badass as long as you don't spend too long thinking about it but has no practical relevance to actual enforcement challenges. It's a
53.
by toraway 5mo ago
The asbestos hypothetical is a bit different than the "bubble popping" economic crisis scenario though. In this world, AI would just continue being adopted and shoved into every nook and cranny into which it can be made to fit, wi
54.
by toraway 5mo ago
Or, it could be like asbestos and the immediate benefits are just too appealing to listen to arguments of skeptical naysayers about some vaguely defined problems that are decades away, if they even happen. I use AI tools daily (because they
55.
by toraway 5mo ago
That reminds me of a chart I saw posted in HN comments recently that someone created tracking bullet points in Claude Code release notes per day that was cited as "proof of a step change" in AI development over the last year. It
56.
by toraway 5mo ago
Despite my own personal preference for static sites, marketing using a CMS under their own control to make content updates seems vastly more reasonable than vibe coding open ended PRs as a codebase they don't understand gradually grows
57.
by toraway 5mo ago
Gemini CLI has improved a lot in the past 6 months or so. Back when I used in the 2.5 Pro era it would get stuck in loops literally like 1/8 conversations and I eventually just gave up despite having access included in my AI Pro plan
58.
by toraway 5mo ago
It may not be bad for the buyers, lenders, or the previous owners who all profit. But even then it could still be bad for regular people/society at large if it incentivizes anti-consumer practices by financial necessity when an otherwi
59.
by toraway 5mo ago
Considering that line is supposed to be written by a young child in-context (who couldn't actually "remember" anything more than a decade earlier, I'm pretty confident the intent was not to reference the actual recent hi
60.
by toraway 5mo ago
> Agents (if implemented well) are an order of magnitude more effective at resolving issues compared to a call centre worker who is reading off a script and churn within 9 months For this to be true, the agent needs to actually b

More ›