hnfong

13 ms·

241.
by hnfong 11mo ago
They do have a valid subtle point though. If we don't think the candle in a simulated universe is a "real candle", why do we consider the intelligence in a simulated universe possibly "real intelligence"? Being a fu
242.
by hnfong 11mo ago
This is basically the Church-Turing thesis and one of the motivations of using tape(paper) and an arbitrary alphabet in the Turing machine model. It's been kinda discussed to oblivion in the last century, interesting that it seems peop
243.
by hnfong 11mo ago
Well, a better retort would be "Human brains are not magic, just physics. Protons, neutrons and electrons don't think". But I think most people get what GP means.
244.
by hnfong 11mo ago
There's a leaderboard that measures user experience, the "lmsys" Chatbot Arena Leaderboard ( https://huggingface.co/spaces/lmarena-ai/lmarena-leaderboard ). Main issue with it these days are that it
245.
by hnfong 11mo ago
Member of ruling class spotted!!
246.
by hnfong 11mo ago
I agree in principle, and I think this is what's happening mostly. But IMHO the public perception of a paper being peer reviewed as somehow "more trustworthy" is also kind of... bad. I mean, being peer reviewed is a signa
247.
by hnfong 11mo ago
Why do you think AI is inherently worse than humans in judging whether a patient has cancer, assuming they are given the same information as the human doctor? Is there some fundamental assumption that makes AI worse, or are you simply proje
248.
by hnfong 11mo ago
The brain does want to predict (in fact I think that's what it's suppose to do), the problem is that the Earth is rather chaotic and unpredictable. If you subject a brain-like neural network to watch planets and stars orbiting aro
249.
by hnfong 11mo ago
Whether people are interchangeable with each other isn't the point. The point is whether AI is interchangeable with jobs currently done by humans. Unless and until AI training requires 1000 different domain experts, the current project
250.
by hnfong 11mo ago
Doesn't matter. The countries with most chaos and internal strife gets a lot of practice fighting wars (civil war). Then the winner of the civil war, who's used to grabbing resources by force, and the one that has perfected war sk
251.
by hnfong 11mo ago
Not seeing any horse heavens, do you have reason to believe humans (i.e. those not among the ruling class) are going to have a different fate from the horses? I agree we can kinda make the argument that abundance is soon upon us, and humani
252.
by hnfong 11mo ago
You decide you need a web server. Ask management chain for approval. Ask IT dept for approval. Ask finance for approval for the expense. Contact Microsoft sales. Buy it. Now you can start developing on it… With open source it’s not just the
253.
by hnfong 11mo ago
there is no flow per se, as it’s a subjective experience. it’s just that our current environment (specifically human environment) is predominantly structured in a way that the past makes an imprint on the present (in terms of biological mem
254.
by hnfong 11mo ago
I don't even think GP knows what negligence is. Generally the law allows people to make mistakes, as long as a reasonable level of care is taken to avoid them (and also you can get away with carelessness if you don't owe any duty
255.
by hnfong 11mo ago
You are totally correct that hallucinated citations do not invalidate the paper. The paper sans citations might be great too (I mean the LLM could generate great stuff, it's possible). But the author(s) of the paper is almost by defini
256.
by hnfong 11mo ago
IMHO what should change is we stop putting "peer reviewed" articles on a pedestal. Even if peer review is as rigorous as code reviewed (the former which is usually unpaid), we all know that reviewed code still has bugs, and a prog
257.
by hnfong 11mo ago
If there was a law against buying the supplies of materials and letting them rot in a storehouse just to deprive competitors of them, your argument would be what OpenAI would try to make in court...
258.
by hnfong 11mo ago
I think you're misreading GP's point, which I think is probably also your point. The comment they replied to was apparently suggesting that everyone struggles in some areas, so it's meaningless to highlight one group's
259.
by hnfong 11mo ago
Saying blue collar work is "low-skilled" is probably inaccurate, I think the important distinction is that most blue collar work requires the person sitting at some physical location doing stuff, while significant amount of white
260.
by hnfong 11mo ago
I see your point, but generally you don't need to roll the dice purely randomly. You generally do get some level of control over your career, your job and the people you work with. With white collar jobs employers are also generally r
261.
by hnfong 11mo ago
Well, it probably correlates to solving leetcode problems within 45 minute interview time limit. /s
262.
by hnfong 11mo ago
Agree that having datestamps on model ids is a good idea, but it's open source, you can download the weights and build on those. In the long run, this is better than the alternative of calling API of a proprietary model and hoping it d
263.
by hnfong 11mo ago
I don't think anyone seriously believes that the Chinese aren't capable, it's more like people believe no matter what happens, USA will still dominate in "high tech" fields. A variant of "American Exceptionalis
264.
by hnfong 11mo ago
Reproducibility of results are also important in some cases. There are consumer-ish hardware that can run large models like DeepSeek 3.x slowly. If you're using LLMs for a specific purpose that is well-served by a particular model, y
265.
by hnfong 11mo ago
If you eliminate ONE job and let's say the job pays $100K, in theory at most $100K goes instead to AI revenue. In practice it's a lot less, nobody is going to move everything to AI if it's just a 10% saving. So, to get a tril
266.
by hnfong 11mo ago
Betting on "growth powered by AI" is the only thing keeping valuations up at this point..
267.
by hnfong 1y ago
Google (sorry) for "James Damore".
268.
by hnfong 1y ago
IIRC, tech in general was trading at very low multiples 10 years ago. Microsoft was also ~ 10, and Google and Facebook were probably < 20 as well.
269.
by hnfong 1y ago
I always suspected Microsoft tried to market it as such otherwise their Windows ME (remember that sh1t?) wouldn't sell...
270.
by hnfong 1y ago
I think Unreal Tournament ran on NT4 as well. Despite Win2k and NT4 kinda having a rep for not for gaming, I found that most games actually did run on them fine. Especially Win2k, probably the most underrated OS of all time in the Windows l

More ›