atq2119

60 ms·

871.
by atq2119 5y ago
This is dangerously careless thinking which completely disregards the complexity of supply chains. Modern IT infrastructure needs lots of different kinds of chips, not just CPUs. Most of them aren't manufactured in the US. If chips sto
872.
by atq2119 5y ago
I've been reading about the Mill for many years now and was fascinated by many of their ideas. However, the more deeply I learn about the details of how CPUs and GPUs work, the more I'm convinced that we're not going to see t
873.
by atq2119 5y ago
I doubt either of us were in the relevant rooms, but the API is very obviously a DX-ification of Nvidia's OptiX (the software-based raytracing solution that they released long before HW raytracing). Furthermore, Nvidia released OpenGL
874.
by atq2119 5y ago
Extensions are the stick with which hardware vendors beat platform companies (in practice, Microsoft) into innovation. That's how we got raytracing, for example.
875.
by atq2119 5y ago
> It’s never about you, until it is. You should assume the government will do the maximum amount of evil given the power they are granted. Add to that: don't only think about what the current government will use those powers for,
876.
by atq2119 5y ago
By user count (as opposed to volume), ransomware is almost certainly a bigger use case of Bitcoin than money laundering and friends.
877.
by atq2119 5y ago
You didn't engage with the core of the argument, which is that the veneer of AI is used to launder the human mistake. Increasing the distance between database input and output gets in the way of human heuristics for recognizing and fix
878.
by atq2119 5y ago
While technically true, the use of tech is still notable here because it was used to launder the mistakes of humans. There is an incorrect tendency to ascribe intelligence to AI, which can make especially laypeople prone to assume that if a
879.
by atq2119 5y ago
My understanding is that the use of neural nets in branch prediction is greatly overstated for marketing purposes. It's really more that different branch predictors are combined into a single one using a simple function (read: linear c
880.
by atq2119 5y ago
It sounds like you do some tracking, but don't do business in Europe. Okay, fine. Do you do only your own tracking? Or do you directly or indirectly sell Europeans' personal data to other companies, who in turn may be doing busine
881.
by atq2119 5y ago
The lithography has obviously been much more salient, and still is, but both are limiting factors. To draw a parallel: as the improvement of computing hardware performance slows down, one would expect more attention to be paid to software p
882.
by atq2119 5y ago
> I have a hard time believing that compute power has been the bottleneck in chip design for many years now. Considering that many of the bread and butter NP-hard problems are actually inspired by chip design, this is blatantly untrue.
883.
by atq2119 5y ago
> If you view GPL code with your browser would that mean that your browser now has to be GPL as well? Some good responses in sibling comments already, but I don't see the narrow answer here, which is: No, because no distribution of
884.
by atq2119 5y ago
Sure. But setting a goal of eliminating it with the purpose of driving elite importance as low as possible seems to me the most ethical way of dealing with the problem of elite overproduction. Additional measures may be useful as well, of c
885.
by atq2119 5y ago
Can't we take the same analysis to drive alternative conclusions instead? For example, why not remove or weaken the reasons why people want to be part of the elite in the first place, perhaps by eliminating it.
886.
by atq2119 5y ago
That seems a bit too simplistic and extreme. More likely, it's because the EU feels too removed from the people. There are plenty of reasons for that, and we're stuck in the pessimum place where democratic reform of the EU which w
887.
by atq2119 5y ago
Yes. Unlike l_\infty, sums of exponentials are smooth everywhere, and smooth functions are easier to optimize. But geometrically, the two can be brought arbitrarily close together by adjusting the exponents with a multiplicative constant.
888.
by atq2119 5y ago
+1 for Multiplicative Weights Update as a technique, and kudos for the nice write-up. Having studied this myself a couple of years ago, I believe an intuitive way to understand its power is that MWU approximately optimizes an l_\infty-norm
889.
by atq2119 5y ago
Even more interesting than GPL, what about AGPL? Perhaps somebody with an AGPL'd project on GitHub and deep enough pockets could force them to open source Copilot... (No, I don't think that's actually plausible, but the scena
890.
by atq2119 5y ago
I don't think we'll see a final node in our lifetimes. Improvements are slowing down and will become a trickle, but that doesn't mean research stops entirely. Consider other mature technology, like the internal combustion eng
891.
by atq2119 5y ago
You do realize that words can have both a legal and a colloquial meaning, right? People say stuff like "he's a fraud" all the time without meaning that in a precise legal sense.
892.
by atq2119 5y ago
Yes, it's the same example, but the answer appears to be subtly different. Acquire and release are only "one-way barriers". Consider the link in the sibling comment by spacechild1.
893.
by atq2119 5y ago
Example: Thread 1 does A.lock, A.unlock, B.lock, B.unlock Thread 2 does B.lock, B.unlock, A.lock, A.unlock That's fine, but swap the middle pair of unlock followed by lock on both threads and you have a standard deadlock based on diffe
894.
by atq2119 5y ago
That reordering would be bad even with different locks because it can introduce deadlocks, as somebody has pointed out. There must be some formal semantics reason which forbids it, and I think the reason boils down to the lock not being a s
895.
by atq2119 5y ago
This stuff is so subtle. It's not actually clear to me right now what part of the memory model would prevent the release store of an earlier unlock from being moved past the acquire load of a later lock. Except that you wouldn't m
896.
by atq2119 5y ago
> That logic would also state that we should no longer call a man a man if he becomes infertile through choice (or otherwise), no? I don't think that's necessarily the case. The argument is less about individual cases and more
897.
by atq2119 5y ago
There is a bimodal distribution of biological traits related to reproduction. Traditionally, women are defined as members of one of those clusters, the one whose members are generally able to give birth during some period of their lives. Th
898.
by atq2119 5y ago
> Is definitely the counter argument I would use - if people have preferred pronouns then what is the ethical argument for ignoring those and using words that would cause them harm So I think a natural conservative argument there would
899.
by atq2119 5y ago
This is the correct answer, plus add loop unrolling etc. It is simply not possible to write an optimizing compiler in which the evaluation order is truly indeterminate yet always the same.
900.
by atq2119 5y ago
> The goal in my mind would be to make a very approachable, reasonable, "safe" reform that's both small and yet makes a difference. Senators get scared of big changes unless the entire political party is on board, which is

More ›