bmh100

14 ms·

121.
by bmh100 10y ago
> So good luck, humanity, in keeping AI under control, should it be developed. This seems to anthropomorphize AI. Why should we expect AI in general to be difficult to control? AlphaGo, while not a strong AI, seems quite harmless.
122.
by bmh100 10y ago
Yes, I found the usage odd. He equates "unpredictable" with "random", which is not how I have usually heard the term in strategic contexts. That means unpredictable, even assuming perfect information. The usual meaning o
123.
by bmh100 10y ago
Would you care to point to or write a blog about the principles of this strategy? It sounds immensely useful to the many C developers out there as a potential set of best practices to adopt.
124.
by bmh100 10y ago
If you are interested in the implications of P vs. NP, as opposed to a survey of the situation around the problem, check out this excellent essay [1] by Scott Aaronson. He goes into interesting consequences about the nature of quantum compu
125.
by bmh100 10y ago
I'm normally an advocate for Fortean for numerical computing, but web servers seem unholy. However, I do appreciate the code reference on querying SQLite from Fortan.
126.
by bmh100 10y ago
After reading the article, my summary of the message is not "backpropagation is a leaky abstraction" but instead "if you don't understand how the derivatives are being calculated, it will come back to bite you". The
127.
by bmh100 10y ago
If I can offer some advice as a former Calc I student, since you are focused on ML, ignore integrals and the fundamental theorem of calculus. Instead, understand the connection between derivatives and antiderivatives and become practiced wi
128.
by bmh100 10y ago
The phrase seems appropriate to me. The students are still working with matrices and linear algebra, focusing on the major algorithms. Going into the element-by-element linear algebra algorithms, for example in matrix multiplication, would
129.
by bmh100 10y ago
I think it is important to keep in mind that theories can sometimes only become testable after sometimes unforeseen advances in technology. Consider the Higgs field, theorized in 1964 [1]. This field was only detected, via the Higgs boson,
130.
by bmh100 10y ago
In the general programming industry over the last several decades, "artificial intelligence" has also applied to simple decision-making algorithms, even as simple as playing checkers. Arthur Samuel [1] was renowned for his work on
131.
by bmh100 10y ago
In case anyone is wondering about what antimatter actually is, antimatter is basically like normal matter, but the electrical charges are reverse. Take hydrogen for example. It is a massive, positive particle (the proton), orbited by a ligh
132.
by bmh100 10y ago
> You will probably come across people saying that the programs are not idiomatic ("enough"). This seems to describe me. I am tempted to rewrite those programs more idiomatically. Thank you for pointing this out.
133.
by bmh100 10y ago
I feel like it's a "right tool for the job" situation. Fortran was literally built for the purpose of numerical computation. Multi-dimensional arrays, non-aliasing arrays and pointers, multi-dimensional array intrinsics, intr
134.
by bmh100 10y ago
What made your work so dependent upon memory bandwidth?
135.
by bmh100 10y ago
How are you using Ruby for physics?
136.
by bmh100 10y ago
Were you using C99 during your HPC programming in academia?
137.
by bmh100 10y ago
It means that the compiler knows that pointers/arrays don't overlap in memory. Data can therefore be safely stored in registers, instead of continually re-read from memory, to account for possible alterations between computations
138.
by bmh100 10y ago
What makes it tricky to use, and where does it fall short?
139.
by bmh100 10y ago
There are some important things to note about these benchmarks. On the Mandelbrot set, I noticed that the program does not use arrays despite the obvious 2D-array structure of the bitmap. Two of the four cores were also underutilized. Next,
140.
by bmh100 10y ago
If you want to get down into the nuts and bolts of numerical computation, with convenient concurrency, you should definitely look into Fortran + OpenMP.
141.
by bmh100 10y ago
I would also like to add that there have been revisions in 2008 and 2015 [1]. [1]: http://fortranwiki.org/fortran/show/Fortran+2015
142.
by bmh100 10y ago
Fortran is an excellent choice for numerical computation. Its non-aliasing assumption allows for compiler optimizations that dramatically speed up computations. It is even easy to get started with concurrent code through OpenMP. When I need
143.
by bmh100 10y ago
It could very well be that way, but what is the evidence for that origin? Why not any number of other plausible explanations? It is not myopic to accept the explanation best supported by available evidence, until sufficient new evidence sup
144.
by bmh100 10y ago
I agree. The article isn't consistent. It references the > Lemon, cocoa, and curry don’t sound great, but it’s not too hard to imagine a decent lemon mole sauce. > Salted cucumbers, sugar, and yogurt might work, but if you count
145.
by bmh100 10y ago
Doesn't that also require at least one additional dimension of spacetime to separate the two universes?
146.
by bmh100 10y ago
Then wouldn't an appropriately constructed sensor be able to record the full linear combination of quantum states?
147.
by bmh100 10y ago
However, doesn't many worlds also necessarily require one or more dimensions for separating the universes from each other?
148.
by bmh100 10y ago
I have a question about wave function collapse, especially tied into many worlds. Let's say that a photon in the double slit experiment could travel straight to a detector, go at a 15 degree angle, or go at a 30 degree angle. Those 3 p
149.
by bmh100 10y ago
Could this connection to the rest of the universe be considered conceptually similar to a form of entanglement?
150.
by bmh100 10y ago
Could you touch on why many worlds is more parsimonious than pilot wave? I'm not sure that I understand that reasoning. Is it because a given "universe" is locally simpler? Would it be accurate to describe the comparison as a

More ›