allenz

30 ms·

241.
by allenz 9y ago
I think that we've made progress. There is now a consensus that those actions were wrong, and should not be repeated.
242.
by allenz 9y ago
It's a trade-off. We lose explainability, but we are able to solve completely new classes of problems.
243.
by allenz 9y ago
No matter how oversubscribed, VM systems have to supply enough servers to meet peak demand, which means inefficiency during the trough. Amazon's spot billing means that they can approach 100% utilization 24/7.
244.
by allenz 9y ago
The edges grow much faster because "molecularly flat regions... have fewer dangling chemical bonds and are thus less favorable attachment sites [for condensing molecules]." Also, snowflakes can take on 3D column shapes depending o
245.
by allenz 9y ago
Snowflakes start from a crystal with hexagonal symmetry and grow via processes that preserve that symmetry. See the top two answers here: https://physics.stackexchange.com/questions/3795/why-are-sno...
246.
by allenz 9y ago
That's a great link, thanks. Journal editors should hire technical editors themselves. This is easier than it sounds since they can just contract with a well-known company at competitive prices. The workflow: first, subject editors fil
247.
by allenz 9y ago
> All the software suggestions to solve this from programmers I agree with you that it is difficult to improve on the peer review process. At the same time, several researchers have pointed out that publishers aren't really involved
248.
by allenz 9y ago
The phrase refers to https://en.wikipedia.org/wiki/Gender_of_connectors_and_faste...
249.
by allenz 9y ago
Both. First, the largest data centers have developed virtualization platforms that are significantly more efficient than vSphere. A major factor is that they provide fine-grained billing for compute, storage, etc instead of selling VMs, whi
250.
by allenz 9y ago
Ok, maybe we can replace the lot of them with a website. But the way I see it, collaboration platforms like Wikipedia, Facebook, and GitHub are valuable, and someone has to maintain the website. I wouldn't agree that this problem i
251.
by allenz 9y ago
You argue that publishing does not need to be as expensive as $2000. Fair enough. I would be happy if we can agree that publishers do something , and that piracy is not an alternative. That point seems to be lost on the majority of comment
252.
by allenz 9y ago
> piracy as a threat to their revenue streams, hence "competition" By your logic, I can "compete" with GM by stealing all of their car inventory and blowing up all their factories in Detroit. I don't think that m
253.
by allenz 9y ago
I agree, editorial committees based on a software workflow could be a real competitor to the current publishers. I'd caution that (1) everything sounds easy from a high level view, when we don't see the complexity, and (2) that th
254.
by allenz 9y ago
Title, together with the domain, is very misleading. The impression is that Wiley is admitting defeat against open access. Instead, the author (not affiliated with Wiley) regrets that piracy is winning over open access.
255.
by allenz 9y ago
By all means, compensate the peer reviewers in addition . But you trivialize what publishers do. The role of the impartial gatekeeper is crucial in our current system. Publishers select the reviewers, analyze the reviews, and accept or rej
256.
by allenz 9y ago
Sure, artists can now sell lessons, but that's not necessarily better than the publisher model. How do you think your analogy would apply to scientific publishing?
257.
by allenz 9y ago
No, and you don't understand the current scientific ecosystem. Researchers are paid for their work through salaries and grants, and therefore benefit from open access. Publishers are not. That's why PLOS must charge authors a fee.
258.
by allenz 9y ago
You argue that publishers are being outcompeted because they don't provide value. First, publishers do provide value.[1] Making copies is a trivial part of publishing; that's a straw man. Publishers add value in their role as gate
259.
by allenz 9y ago
In theory, an IDE should be able to toggle between a literate programming layout and the actual structure that gets executed. Both views are valuable. If we take your argument to extremes, all programmers should write in assembly. Clearly p
260.
by allenz 9y ago
The entire series is great: http://www.nybooks.com/topics/on-consciousness/ . If you're interested, there are several high-quality popular philosophy podcasts as well: https://www.reddit.com/r&
261.
by allenz 9y ago
In theory, yes. In reality, we don't have enough data for that to work. A model that takes the whole genome as input is excessively expressive and would overfit, finding spurious correlations everywhere. In the near term, we still need
262.
by allenz 9y ago
While each little piece isn't very impressive, taken together I think it makes life significantly more convenient, anticipating your every need. The promise of IoT is that everything "just works", like magic. It's like h
263.
by allenz 9y ago
You can distinguish artisan sushi from mass produced sushi if you pay attention, since they're made in very different ways. Artisan sushi is airier with softer rice. You might be interested in these analyses: http://boingboi
264.
by allenz 9y ago
It's not that easy because computers can't read most data. How would a computer know whether an image/video file contains an encrypted message? https://en.wikipedia.org/wiki/Steganography
265.
by allenz 9y ago
I don't feel that ambition and risk is a bad thing. Elon Musk is crazier than you give him credit for. He started Neuralink, which is even more of a moonshot biotech company. I think that starting a new car/rocket company was risk
266.
by allenz 9y ago
It works for arbitrary data because > a hash function is applied to each element in the original multiset to obtain a multiset of uniformly distributed random numbers with the same cardinality as the original multiset
267.
by allenz 9y ago
There is also some discussion of these posts at https://www.reddit.com/r/AMADisasters/
268.
by allenz 9y ago
Neural networks don't do anything as sophisticated as self-referential introspection. They just fit the outputs you train them with. The training data you provide would have to include the desired explanations. Consistency is enforced
269.
by allenz 9y ago
It's theoretically possible for a neural net to do this; the network just needs to have the explanation as an output. I agree that decision trees would be more reliable and easier to train, but I'm not sure if hardcoding every fea
270.
by allenz 9y ago
Edge is built into Windows at a lower level than UWP. It is UWP-like in the sense that it has the modern app UX on all Microsoft products, it uses the new APPX package format, it favors the new WinRT API, it uses the new AppContainer sandbo

More ›