SideQuark

6 ms·

1.
by SideQuark 5d ago
Laptops are already counted in PC sales, since they are PCs.
2.
by SideQuark 5d ago
Games doubling in price hides less copies bought, much like movies selling far less tickets yet having higher revenues. No matter how you count it, PC gaming has been in decline for a long time.
3.
by SideQuark 6d ago
PC sales peaked 15 years ago, and have been declining ever since. Most of the best selling PC games are from 15-30 years ago.
4.
by SideQuark 6d ago
As someone making mobile apps used by billions, demands from mobile users certainly grow and want this class of hardware capability. We spend significant effort trying to figure out how to deliver features asked for by our users for high en
5.
by SideQuark 6d ago
Pretty much all cutting edge graphics papers now use some form of NN. To ignore the techniques and gains at the cutting edge would leave any hardware mfr behind. It’s not nonsense or marketing speak. It’s genuine capability improving techno
6.
by SideQuark 6d ago
Datacenters don’t run on desktop class GPUs. They’re well past that. It’s why NVidia moved most production of desktop GPUs to data center class hardware.
7.
by SideQuark 17d ago
That’s completely backwards by orders of magnitude. Modern CPUs outrun L1 cache speed by a lot, L1 cache speed is on order of 1000-5000 GB/s, a GB/s is real bits used, is 8x of Gbps, and network speeds include significant overhead
8.
by SideQuark 17d ago
The vast majority of all commits to Linux were paid for by big business (around 85% last year, prev years similar). Intel, Google, AMD, IBM, Meta…. https://commandlinux.com/statistics/linux-kernel-contributor...
9.
by SideQuark 20d ago
Not sure what you’re intending to show. That page shows 99.5%+ uptime across all 11 of the services provided. To do that at a small company world cost significant money. To have it integrated and cheap for all those small shops is very valu
10.
by SideQuark 23d ago
EU law requires the mark. Did you know about the level of the “so many levels”? If such a large market requires behavior by law, and quite reasonably so, it’s also quite reasonable to mark AI generated content as such. All the slippery slop
11.
by SideQuark 23d ago
If you read the article instead of the headline…….. it’s adding a fairly standard mark to AI generated images to let’s others know, in the same manner a giant swath of the GebAI industry has agreed to.
12.
by SideQuark 25d ago
Agreed. But there are reasons it is perfect legal for a library to digitize physical copies they own, and then can lend, also under the legal name of "Controlled digital lending". Clearly public domain works are fine, and US law a
13.
by SideQuark 26d ago
> suit is mainly over whether controlled digital lending is legal at all No, it was not, even supported by the quotes you pulled. Libraries right now, with publisher blessing, offer all manner of controlled digital lending. The suit was
14.
by SideQuark 26d ago
Yeah, that worked really well for captchas. Now even simple models far surpass humans at recognition. So there’s plenty of evidence. Now provide evidence it’s technically possible to make any font “ that is financially infeasible for LLMs t
15.
by SideQuark 26d ago
> I'd love to hear your thoughts on that. “Claude: make the scraper mimic a screen reader.” And just like that, in 10 seconds, their site feeds my “screen reader” the real words.
16.
by SideQuark 26d ago
So jealousy is the argument? If someone reads my work am I worse off? Why would I post it to the web? If someone indexes it so others can find it am I worse off? Should I complain they also make money helping others find it? If someone buil
17.
by SideQuark 27d ago
You misrepresent the non cs ones by a factor of 5-10, even on cpu. First, you’re calling the prng twice per aes single call. That seems pretty dishonest already. And using Golang? That ludicrously slow also. To tell us what a cpu can do, do
18.
by SideQuark 28d ago
Same - very good article.
19.
by SideQuark 28d ago
As others here have pointed out, this is nonsense. The vast majority of PRNG calls on the planet are extremely high perf simulations, where crypto secure versions are a ludicrous cost in speed, energy, and sheer stupidity. That you and othe
20.
by SideQuark 1mo ago
Those were leading capital letters words. This is ALL CAPITAL LETTERS, which is different. As others pointed out, it’s used extensively in the click bait arena.
21.
by SideQuark 1mo ago
> The UC system and many other of the top universities in the country have also stopped looking at SAT scores. Which led to a spectacular failure of incoming students to do basic math, lead to massive increases in failures in first year
22.
by SideQuark 1mo ago
Courts have ruled many things besides encryption count as anti-copying measures, and that the laws apply retroactively, so it’s not so clear such patches even on 360 are legal. Try selling such patches and test it.
23.
by SideQuark 1mo ago
When you conflate an 8-bit xor cipher with AES and a 10 line python NN with the complexity of trying to break AES with a NN by having "the model ... predict the key", I can see how you think such a thing is possible. It's non
24.
by SideQuark 1mo ago
Employing anti-protection methods to do to has been illegal since 1996 WIPO treaties, and then much more strongly in the 2002 DMCA. So around 30 years now.
25.
by SideQuark 1mo ago
Have you ever built a NN model? Have you ever broken a crypto system, even a small one?
26.
by SideQuark 1mo ago
There are plenty of algorithms that quantum algorithms give no benefit to. Quantum is NOT a free for all making everything faster. It’s only faster at two very,very restricted things: hidden subgroup problem (which is used for RSA, but most
27.
by SideQuark 2mo ago
You’re not giving it to a banker, you’re trading risk for return and flexibility. One can access savings at any time, any amount. Not true with bonds, maybe if you fiddle with indices. Also bond returns have averaged 5% over decades, not 7.
28.
by SideQuark 2mo ago
Single event upsets are forced on you by physics. There is no possible way to stop them all.
29.
by SideQuark 2mo ago
> Put all of these trends together, and I think that in 10-15 years, we are going to have consumer PCs (and phones!) running models doing pretty much anything that frontier models can do right now. Not likely. The last 50 years had Moore
30.
by SideQuark 2mo ago
That's not recursion. And reversing arrows on this category turns base cases into co-base, or start cases, (and these are still often called base cases in the literature), so it is still required. Reversing arrows again, back to recurs

More ›