cubefox

22 ms·

691.
by cubefox 7mo ago
The geometric mean seems to be the natural mean for relative comparisons between lengths, because the mean of (Planck length, observable universe) is clearly very different from the mean of (house, observable universe).
692.
by cubefox 7mo ago
> Not on a free app 30% of 0 is also 0. They are already cross-subsidizing it. > I don't think Apple should be coerced to host content for free and cross-subsidize it from other paid content. Nobody said they should be.
693.
by cubefox 7mo ago
She: It's not that big. He: I think we can agree everything below the average between a Planck length and the size of the observable universe is objectively small, and everything above is objectively large. Using the geometric mean, th
694.
by cubefox 7mo ago
> Only if you charge for your app - and how much free labor and bandwidth do you give away? Apple gives away millions. I guess a ~2% fee would cover those costs.
695.
by cubefox 7mo ago
I think you mean Wero
696.
by cubefox 7mo ago
And only 30% fees, just for being on the app store!
697.
by cubefox 7mo ago
> Intel in 2019 had $72 billion in revenue and 110k employees. When he was fired in 2024, Intel had $54 billion in revenue and 124k employees. He also didn't cut dividend until 2024. He nearly put Intel into the grave in 3 years. Th
698.
by cubefox 7mo ago
Do you believe heroin is immoral? I don't. I think it's dangerous, which is bad, and it causes addiction, which reduces freedom more than banning it.
699.
by cubefox 7mo ago
> I think the real mistake was Apple allowing to enter a non-ASCII password in the first place. No that's obviously crazy!
700.
by cubefox 7mo ago
> Illegal stills make quality control impossible, so legalisation and government certified testing can make it safe. Another way to increase safety is to reduce the availability of illegal stills without quality control by enforcing the
701.
by cubefox 7mo ago
> Half the microplastics in our body is from cars I doubt it. The article suggests most is from inhaling indoor dust from synthetic textiles.
702.
by cubefox 7mo ago
They talk about three applications for the coming cooperative vectors / linear algebra API: Neural Materials, Neural Radiance Caching, Neural Texture Compression. But what about already much more popular ML-based applications: upscalin
703.
by cubefox 7mo ago
The article says this is probably wrong. We breathe in much more.
704.
by cubefox 7mo ago
That's the silliest LLM comment I saw in a while.
705.
by cubefox 7mo ago
> It becomes a larger problem as the world moves away from fossil fuels like natural gas. I actually remember a similar problem from some compound that was mainly formed as a byproduct of some old Canadian nuclear reactor design. As the
706.
by cubefox 7mo ago
The article briefly touches on insufficient recycling. Though it's not clear for which applications helium recycling is technically/economically feasible and for which it isn't.
707.
by cubefox 7mo ago
The reason helium can't be produced chemically (like hydrogen can be produced e.g. from water) is that there are no natural chemical compounds which contain helium. That's because it doesn't form those compounds in the first
708.
by cubefox 7mo ago
The radiation in space is not "small dose". It's firmly established that it is quite bad for human health.
709.
by cubefox 7mo ago
You should read the article. The problem is the destructive effect of radiation on the human body, not on the space hardware.
710.
by cubefox 7mo ago
> If you are desperate for extra safety then just include multiple computers, literally what spacex does. That does nothing to protect the human body from the radiation damage. Talking about microchips is a distraction.
711.
by cubefox 7mo ago
> The math doesn't need a 'true' or 'false' distribution; that just falls out of the use of a model ('false') to approximate reality ('true'). Looks like a contradiction. If you identify reali
712.
by cubefox 7mo ago
It's far worse not to have sufficient electricity during the night or on overcast days. You can just increase nuclear electricity prices during that time to make up for the lost revenue from sunny days.
713.
by cubefox 7mo ago
Creating the Futurescape for the Fifth Element (1997) (2019)
714.
by cubefox 7mo ago
I think the film would have been better (though perhaps less successful) if Besson had toned down the occasionally exaggerated tomfoolery, like Chris Tucker's character, or the spaceship Evil (the orb described in the article) which fe
715.
by cubefox 7mo ago
> To the first point, I think that the KL divergence is indeed symmetric in this case, 0.4 * ln(0.4 / 0.6) + 0.6 * ln(0.6 / 0.4) no matter which direction you go. But my argument also works for any other probability distributio
716.
by cubefox 7mo ago
You misunderstood what I was saying. I was not suggesting that the KL divergence is symmetric. I was saying that it would be symmetric ( and independent of the "truth" of a distribution) if it was interpreted as the quoted me
717.
by cubefox 7mo ago
> D(P||Q) = measure of how much our model Q differs from the true distribution P. In other words, we care about how much P and Q differ from each other in the world where P is true, which explains why KL-div is not symmetric. I don&#x2
718.
by cubefox 7mo ago
Unfortunately all these intuitions rely on a distinction between a "true" distribution P and a "false" distribution Q. So they don't work for a subjective probability interpretation where it doesn't make sense
719.
by cubefox 7mo ago
They are: https://sheets.works/data-viz/hire
720.
by cubefox 7mo ago
Yes. They are likely also advertising for themselves with how viral their ads are. The article is featured on their website.

More ›