hnfong

7 ms·

91.
by hnfong 4mo ago
This has changed (in a nit-picky way) - Gemini is now generally available to the public in Hong Kong. ChatGPT and Claude are not available. Generally my impression is that OpenAI isn't that anal about service providers reselling ChatGP
92.
by hnfong 4mo ago
There are so many Chinese open weights models that any company with resources can run them in-house (or with a trusted provider). There might be some valid concerns about model alignment, but at least the model running in-house isn't g
93.
by hnfong 4mo ago
You're not alone. That's essentially the experience of most people having English as a second language.
94.
by hnfong 4mo ago
Not sure where you got the "I don't want to take any responsibility for being aware of my surroundings". GP simply pointed out cyclists are apparently super unfriendly to deaf people, inferred from the experience where GP mad
95.
by hnfong 4mo ago
You can disagree with the judgment as long as you follow the orders.
96.
by hnfong 4mo ago
You might be factually correct about the privileged part, but if you think about it, it's simply a privilege to not be a slave of capitalism...
97.
by hnfong 4mo ago
If the countries were reversed, and some Chinese software implemented an equivalent "security feature" to track US users, it would be all over the news about how China is conducting spying and espionage on America. Or maybe you do
98.
by hnfong 4mo ago
I still don't understand why you don't just wake up earlier. It's not like without DST you have to work so late that you don't have enough hours for sleep, right?
99.
by hnfong 4mo ago
That's simply not true. ~70B models can run fine (albeit somewhat slow) on consumer hardware with 64GB RAM. There are heavily quantized (Q1.x) models that are still usable on similar hardware. Granted recently there haven't been a
100.
by hnfong 4mo ago
Hey guys... I have a wonderful idea. ... Let's write ECMAScript for portability.
101.
by hnfong 4mo ago
... yes? Also, what if there were malicious submissions? Then they would definitely go to court. How can you tell?
102.
by hnfong 4mo ago
Also, in most countries (especially the US), also a defamation problem.
103.
by hnfong 4mo ago
Qwen models are super useful for those running local. And there are valid reasons to run local, even if performance (quality and speed) aren't best.
104.
by hnfong 4mo ago
It's a proxy for what you actually want to measure. Note that after the model generated a bunch of (intermediary) code, they still have to have it tested and get bugs fixed (via the agent/harness). In this "one shot" you
105.
by hnfong 4mo ago
There was some funny suggestion online with using Classical Chinese (which has a similar status to Latin in Europe, and it uses at least 50% less characters, probably similar savings with tokens) to reason. Don't know whether the rea
106.
by hnfong 4mo ago
> I'm just waiting for some corporate IP law firm to learn about Huggingface. Presumably they already know. The issue is that IP law firms are tiny compared to the trillions of capital pouring into "AI". And if you belie
107.
by hnfong 4mo ago
> The man is smart but IMO shouldn’t be running the company- he should be a CTO and let a business person make the decisions. Well he didn't like Sam doing it, and whoever CEO puppet he finds will be a Sam-like figure anyway.
108.
by hnfong 4mo ago
Jury's still out whether they refuse to do that. Currently it's just that they can't immediately implement those checks. They'd need some biometric check on API calls and some magical way to prevent the authenticated per
109.
by hnfong 4mo ago
The weirdest time was when I got on a girl sorority email list. Told them they got the wrong email a couple times, gave up, and just added a mail filter...
110.
by hnfong 4mo ago
The ruthlessness of the Microsoft that we remember was probably due to influence from Bill Gates. He is famous for hyper-competitiveness and strong desire to win at all costs.
111.
by hnfong 4mo ago
GP's argument is essentially this: https://en.wikipedia.org/wiki/Evil_demon Perhaps you might think this is bullshit because * obviously* this world is real and not an illusion and there is * obviously* no evil
112.
by hnfong 4mo ago
Don't know why you're downvoted so much, but your observation is spot on. This is essentially Descartes evil demon issue. If you can't disprove that an evil demon (with god-level powers) is deceiving you at everything you per
113.
by hnfong 5mo ago
I have no idea why anyone (especially those here) would be dismissive of genAI from ChatGPT(2022) onwards. It was obviously a new tech, and was obviously good enough that more resources would be invested to improve it, and it really amazes
114.
by hnfong 5mo ago
It might reflect a prejudice of modern human society to think more highly of people who are intelligent, and think less of people who we deem stupid. I'd say AI should not by default gain any social status as human-equivalent even if t
115.
by hnfong 5mo ago
My point is that the scientific establishment is already basically like those boards that can limit your ability to do research unless you have a license. The "license" is a PhD (from a reputable institution) and publications in
116.
by hnfong 5mo ago
I think you misunderstand. You think blue zones are "thoroughly debunked". And it might be even true that blue zones aren't real. But I am not seeing any evidence that the general public is aware of this. And thus we need mor
117.
by hnfong 5mo ago
This is your personal opinion or general consensus?
118.
by hnfong 5mo ago
I can't tell if you think academia is "enforcing existing knowledge and consensus" or "preserving intellectual freedom"... My personal take is that academia is doing very well on the former, and not much on the latt
119.
by hnfong 5mo ago
The prompt is literally "why is the sky blue?" and consists of 7 tokens. It's probably too small for the timings to be taken seriously.
120.
by hnfong 5mo ago
What do you mean? You pretty much described classic moral panic.

More ›