ronsor

4 ms·

31.
by ronsor 1mo ago
"Congress shall make no law [...] abridging the freedom of speech" We've been down this road before with cryptography. No, it's not going to work.
32.
by ronsor 1mo ago
I couldn't even get Fable to build my auth endpoints at all!
33.
by ronsor 1mo ago
Fable is way too expensive for basic software stuff. Other models are more than good enough for that. In general, if Fable isn't blocking you, there's a high chance a lower tier model would work fine.
34.
by ronsor 1mo ago
Coincidentally the last time native apps were good.
35.
by ronsor 1mo ago
Even Brussels knows that would be futile
36.
by ronsor 1mo ago
Before I read this, what is there about C3 [0] you don't like? [0] https://c3-lang.org
37.
by ronsor 1mo ago
I don't think Spain or Sweden are very responsible for most of the Middle Eastern woes. If anything, blame the US.
38.
by ronsor 1mo ago
Europe has learned the lesson that extremists are bad, but it has not learned the lesson that the tools shouldn't exist.
39.
by ronsor 1mo ago
I used to think OpenAI was the only bad actor until I realized they were the only honestly corporate actor. The other AI companies obfuscate their goals much better.
40.
by ronsor 1mo ago
Exactly. Astra will likely get its Mythos moment, and then another Chinese lab releases its equivalent three months later while Anthropic releases a newer, better model.
41.
by ronsor 1mo ago
You may be safe from a catastrophe, but you'll never be safe from the wildlife.
42.
by ronsor 1mo ago
So it can hack the cellular network instead?
43.
by ronsor 2mo ago
This is literally already happening and has been happening for over half a decade now: https://www.datacenterdynamics.com/en/analysis/data-centers-...
44.
by ronsor 2mo ago
Shouldn't this already be covered by the Fourth and Fifth Amendments?
45.
by ronsor 2mo ago
They want AI companies to shut down. Unfortunately, "we simply want you to not exist" is not realistically actionable or even compatible with a free society.
46.
by ronsor 2mo ago
You just discovered that principles are rare. Now let me show you my preferences of convenience. "Information wants to be free as long as I benefit from not paying."
47.
by ronsor 2mo ago
> Much of humanity's progress has involved the abolishment of immoral laws. So you are saying to abolish copyright?
48.
by ronsor 2mo ago
You'll need the Person-to-Prompt Training Protocol to go with it.
49.
by ronsor 2mo ago
They are literally not allowed to publish those scans. Half the reason the books get trashed in this process is because the first sale doctrine keeps copyright from strangling all the freedom in this narrow area.
50.
by ronsor 2mo ago
I think Stripe does enough already. Fraud detection is a crapshoot regardless; I've had legitimate transactions rejected at random without explanation.
51.
by ronsor 2mo ago
The one thing I trust is that Chinese prices aren't overinflated. They're almost certainly closer to the actual cost of inference + training amortization than what Western labs are offering. DeepSeek's price hike is mostly dr
52.
by ronsor 2mo ago
I don't know why cartels and oligarchs are always brought out as a reason for this stuff. One, banks do KYC anyway. Two, cartels and oligarchs will pivot to outright fraud overnight; reporting rules don't discourage them.
53.
by ronsor 2mo ago
Implicit declarations are a classic! Current GCC and clang error on them by default now, but that's a rather recent change.
54.
by ronsor 2mo ago
We should do this for taxi rides and cars.
55.
by ronsor 2mo ago
This is Hacker News. You know people don't follow links and read.
56.
by ronsor 2mo ago
People don't focus on improving digital ownership because that requires butchering the sacred cow of copyright and DRM. These are the root of the problems and no one wants to touch them.
57.
by ronsor 2mo ago
At <50M parameters, training costs are completely trivial. You'll spend a lot more on your rent this month.
58.
by ronsor 2mo ago
I disagree. The parents are the ones who should be supervising. The work should not be pushed off to OSS devs or random website operators. Facebook is bad, but I'd rather deal with one Facebook than kill 1000 normal sites with bad legi
59.
by ronsor 2mo ago
Don't be anything?
60.
by ronsor 2mo ago
Doesn't matter. Even by the typical robots.txt definition, a bot has to operate more or less fully autonomously. If the user initiated the request, it doesn't count, and it shouldn't follow robots.txt.

More ›