HN user

vitalyan8184

-1 karma
Posts0
Comments16
View on HN
No posts found.

[citation needed]

common sense? businesses like making money and don't like spending money. why build glorified warehouses where it isn't economical to do so?

And, AIUI, the only reason they use up any water, rather than using a completely closed-loop system, is because it's cheaper that way.

evaporative cooling takes way less power to operate.

And it’s a non uniform tightly clustered spread

yes, there probably aren't data centers in the Death Valley and Mojave. they're being built where water is cheap. water is cheap where it's abundant.

That’s a lot.

an 1/400th water usage increase is not the catastrophe the public is being manipulated into believing.

hello, stalker! a throwaway is not a "troll account", it's an account dedicated to expressing opinions you preemptively know the local commissars are going to downvote and flag.

for example, consider this comment https://news.ycombinator.com/item?id=48945532. did that polite reminder that the tan suit was not the only controversial thing about President Obama deserve -4 points and [flagged]? or this https://news.ycombinator.com/item?id=48643531. did that simple factual statement that "fiat can be taken away from you" deserve -4? I could go on and on and on. few things I post would be controversial outside of terminally online spaces.

Which explains one of the greatest social paradoxes of today: approximately everyone is using AI and almost everyone is hating it. I don't mean HN, but public opinion polling for the tech is absolutely awful.

the legacy media is running an unapologetically dishonest atrocity propaganda campaign against it. just consider how much we hear about AI guzzling water, while in reality all US data centers account for 0.25% of US water usage and dedicated AI data centers account for a fraction of that. the rest of their arguments are similarly deceiving.

In the US, it had an approval level comparable to Iran.

Iran had also been a target of atrocity propaganda campaign back in January when CIA and Mossad were preparing the American public for the imminent war.

Qwen 3.8 3 days ago

their "genuine moat" is that mythos is the only super heavyweight model right now. it's always been possible to train a 10T model and get 10% more performance over a 1T model.

had mythos been just Opus 5, with the same size and price as the previous opuses, then yeah, that would be a tie-breaker. but it's not.

Qwen 3.8 3 days ago

nvidia and amd don't give a flying fuck about end users right now while they can milk triple digit markups from infinite money VCs via data center GPUs.

again, you're ascribing human level of consciousness to much simpler creatures driven primarily by instinct.

and you are aware that even humans, with our much longer lifespan and much higher cognition, can endure living in captivity for years just fine, right? hell, even outside of captivity, a double digit percentage of us doesn't have a single room of living space, and the only outside most of us see is covered with asphalt and concrete.

in civilized countries, industrially farmed animals live short, uneventful lives devoid of hunger and pain, injury and disease, heat and cold, then die instant, painless deaths. in the wild, the kind of animals we farm rarely get to die of old age - most end up being mauled and eaten alive by predators.

their lives are only bad if you assume they possess adult human level of conscience, which they obviously don't.

with open weight models, you have three other options

A) use a provider that pinky-swears not to store your data. they obviously don't give a fuck about 'distillation attacks', so they have little motivation to voluntarily monitor and store your queries. reasonably high likelihood of privacy.

B) rent the hardware and run the model yourself. very high likelihood of privacy.

C) buy the hardware and run the model yourself. absolute certainty of privacy.