murbard2

11 ms·

61.
by murbard2 6y ago
A fun fact about monster trucks is that they do exactly that: spin they massive wheels while in the air to control their angle during jumps.
62.
by murbard2 6y ago
How would this method handle the bar I posted?
63.
by murbard2 6y ago
A fun example is this bar from Liebenstraum no 3, https://imgur.com/a/ud1VlJF It's the second bar on, the left hand goes A F# B# D# B# F# - Aflat F# C Eflat C F# Only the first note has actually changed though. E
64.
by murbard2 6y ago
> This will not teach you to play piano. I agree, but tools like this or Synthesia can potentially help you practice the piano, which is what you should be spending most of your time on.
65.
by murbard2 6y ago
Musescore does do a great job and is open source, so there's that. Another option is to have users manually upload midi with a proper score attached to some store or community repo with user reviews. It's less satisfying as a solu
66.
by murbard2 6y ago
It's funny, I've had almost the same experience: decided to start playing again during Covid, and wanted to use the midi port to gamify things a bit. I've been using Synthesia[1] on an iPad, which also lets you take an arbitr
67.
by murbard2 6y ago
Are advertisement cookies now considered "stalking"? The news media does engage in _actual_ stalking for stories. And no, Facebook doesn't actually sell user data, no matter how much their competitors like to repeat it. Faceb
68.
by murbard2 6y ago
People don't use Facebook against their will but many are forced to comply to GDRP even though they oppose it. Your outrage is misplaced. The press competes with Facebook, they have a direct incentive to paint them as a Goliath fightin
69.
by murbard2 6y ago
Great product placement.
70.
by murbard2 6y ago
What's in a title? "Paris hails victory". No. The city government of Paris hails a victory. Whether that's a victory for the city and its inhabitants is a very different matter.
71.
by murbard2 6y ago
Your comment seems like it would rule out a photon rocket , which proves too much. I think the resolution is that the faster you go, the heavier you are and the more thrust you need to increase your dv.
72.
by murbard2 6y ago
That's the name of the legal entity. The actual name, if memory serves, is "X, the moonshot factory".
73.
by murbard2 6y ago
> There is absolutely no training signal that tells an RNN to remember something beyond the BPTT horizon. So why would it? Because they generalize. Char-RNN learn to balance parenthesis separated by a longer distance than the BPTT window
74.
by murbard2 6y ago
You need to encode the delta between your model and what you observe, and that delta will be expensive to encode if your model over-generalizes.
75.
by murbard2 6y ago
I also found it very baffling from a French standpoint. Obviously, spouting out des "gros mots" in polite society is frowned upon but the the magical thinking around 4 letter words in the US is on a whole other level. The country
76.
by murbard2 6y ago
Over-generalizing means that you will compress worse, as does under-generalizing. Can't use the mere fact that "compression is going on", you have to look into utility, costs, etc.
77.
by murbard2 6y ago
Bastiat is specifically referring to rent seeking, occupational licensing, protectionism, etc. The Bastiat quote is on point if your criticism of Purdue Pharma is that it profited from the medical licensing laws and drug prohibition laws th
78.
by murbard2 6y ago
It's cringe, but I don't find it hard to read, anymore than knowning numbers precludes me from reading "4c7u411y p0274813 3x3cu74813".
79.
by murbard2 6y ago
Many of the modern ones are based on deep learning and might indeed capture priors for faces. However, other approaches do not depend on machine-learning. The priors that they implicitly use are weak or rather general (e.g. natural images h
80.
by murbard2 6y ago
Lots of neural-network approaches for this which work quite well but are also likely to learn priors from the data which can introduce tricky issues for police work. There are however approaches which aren't based on machine learning a
81.
by murbard2 6y ago
Before casting stones, have they checked with Ryan Gosling to make sure that wasn't him behind a giant magnifier at that window?
82.
by murbard2 6y ago
Not quite. Super-resolution from face prior is probably a bad idea if you're looking to identify someone, but super-resolution from video works. It doesn't make up information based on what it expects the face to be, but collect
83.
by murbard2 6y ago
There's a nonce on each account. If you see a transaction from the same account with the same nonce it's a double spend . Nonce can't be skipped either, so a double spend requires two transactions with the same nonce.
84.
by murbard2 7y ago
I don't think it is about political pressure in the sense that, I doubt the USG fundamentally cares about whether or not Iranians can use GitHub or not. For companies like Google or Microsoft, this is about minimising the attack surfac
85.
by murbard2 7y ago
I think the point will be selective enforcement. Make a completely unreasonable request that travelers are unlikely to perfectly comply with (even if inadvertently -- do you remember your Digg profile?) and later enforce a penalty at your d
86.
by murbard2 7y ago
The US is forcing its tourists to give away a list of all their social media accounts and all their email accounts. If you're a foreign journalist writing pseudonymously for your safety, you must now share that information with the US
87.
by murbard2 7y ago
Sadly, this probably beats dealing with the bs they would have received from the USG.
88.
by murbard2 8y ago
The program collects statistics based on your previous 5 key-presses. That's one way to do it, but there's a much better, extremely elegant algorithm for quickly making predictions by averaging over a doubly exponential number of
89.
by murbard2 8y ago
For that, you can replace Kolmogorov complexity with Levin complexity, which penalizes the Kolmogorov complexity by adding the log of the execution time. Since sufficiently advanced Turing machines are all equivalent up to a multiplicative
90.
by murbard2 8y ago
You can go Big endian all the way, [https] [com] [phishing] [com] [example] [foo.html]

More ›