YorkshireSeason

9 ms·

31.
by YorkshireSeason 4y ago
No, this is October 2022. It is about Schoenbohm at the BSI, not Maassen and the BND!
32.
by YorkshireSeason 4y ago
Rumours are that he has been consistently against backdooring crypto, and that is why he's not well-liked in government circles: https://www.heise.de/meinung/Kommentar-BSI-Affaere-Boehmerma... I'm not in a po
33.
by YorkshireSeason 4y ago
Deans typically lack detailed technical knowledge to evaluate candidates (not to mention time), they just go with the flow, and want to run the department smoothly. I think the real mechanism is this: those who make the decision (those with
34.
by YorkshireSeason 4y ago
I would imagine that it's the metrics universities and funding agencies apply in promotion decisions. For example, my (well-known) university decided to measure impact, and takes "Twitter engagement" as one proxy metric for
35.
by YorkshireSeason 4y ago
What scientifically credible, i.e. at least stably reproducible, connection does the head twitch response in mice have to do with depression in humans?
36.
by YorkshireSeason 4y ago
There is nothing credible. This research direction is an example of "cargo-cult science" [1] as Feynman understands the term. (That does not mean that the hypothesis is necessarily wrong.) There is a reason for "in mice&quo
37.
by YorkshireSeason 4y ago
It is not the case that "the search space is just the training data" , the search space in NNs are some variants of real multivariate functions. The training data can be all manner of things, in LLMs, they are text in natural la
38.
by YorkshireSeason 4y ago
Modern DL uses stochastic gradient descent, which Domingos's paper doesn't deal with, in fact he leaves the generalisation as an open problem. Be that as it may, universality in computing models is a well known issue. I could al
39.
by YorkshireSeason 4y ago
How do you explain the tangible progress in ML?
40.
by YorkshireSeason 4y ago
I don't understand those objections. Yes, one of the key insights of modern DL-based ML is that size matters, neural networks start producing interesting generalisation only once you have large parameter spaces. Indeed, one might wonde
41.
by YorkshireSeason 4y ago
This sounds like you consider what has been called "feature engineering" in older ML research as being fundamental to modern ML successes. But the trajectory has been the opposite direction, away from feature engineering, and in
42.
by YorkshireSeason 4y ago
When Newton described the laws of gravity, now often referred to as Newtonian physics, did he explain what gravity was, in other words, did he have a "theory of operation of the model, except the implicit one"? No, au contraire !
43.
by YorkshireSeason 4y ago
The alternative is the come to terms with one's sexuality, just like everybody else. In other words, in most relevant cases here, to come out as a gay man.
44.
by YorkshireSeason 4y ago
First paper to implement NNs on GPUs was [1]. Schmidhuber's survey [2] also mentions it. [1] K.-S. Oh, K. Jung, GPU implementation of neural networks. https://koreascience.kr/article/CFKO200725752349451.pdf [2]
45.
by YorkshireSeason 4y ago
I found myself in a somewhat similar situation. I learned to thing long-term: (1) Publish initial ideas not in top conferences, but in second rate venues. Most publications in top conferences are, clearly, refinements of ideas that were fir
46.
by YorkshireSeason 5y ago
Could you elaborate on the connection between Deep Learning and type theory, resp. topos theory that you allude to? It sounds fascinating. At the some time, I'm slighly surprised that I am not aware of results in this direction.
47.
by YorkshireSeason 5y ago
You are entirely correct, those countries have good reason to fear Russia (with the additional complexity of substantial numbers of ethnic Russians in those countries), but, as I wrote a month ago [1], there is a symmetry here and Russia al
48.
by YorkshireSeason 5y ago
"You forgot Poland" [1] Ukraine and Poland both invaded Irak in a recent war of aggression [2, 3]. Have you contemplated the question of helping to bring to trial the Polish leaders that signed off this violation of the internat
49.
by YorkshireSeason 5y ago
> strong correlation between rent prices and fertility rates. I assume you mean a positive correlation. When the Soviet Union collapsed, many of the former Eastern Block countries experienced both, a reduction in fertility and too mu
50.
by YorkshireSeason 5y ago
As "baq" also says: you pay Arm (or Apple, or Qualcomm or ...) for the micro-architecture. I expect Arm to drop licensing costs for their ISA to zero to compete with RISC-V at the low end. Right now, the Arm ISA has much better mi
51.
by YorkshireSeason 5y ago
> real value will be in the implementations themselves. It is not that difficult to re-target a micro-architecture (what you call implementation) to a new ISA, especially if the ISAs relatively close, as most modern ISAs are. RISC-V an
52.
by YorkshireSeason 5y ago
> Intel has realized this One of the reasons Intel is spending money on RISC-V could be that they want to undercut Arm, as the latter are now beginning to be competitive in Intel's X86 world. RISC-V is going to succeed at all as a
53.
by YorkshireSeason 5y ago
Putin as been in power since 1999. In that time NATO countries have invaded (direcly or through proxy forces): - Iraq - Syria - Yemen - Libya - Afghanistan Just two days ago NATO countries conducted drone strikes in Somalia. NATO aligned f
54.
by YorkshireSeason 5y ago
Have you got substantial comments about the content of the article, the relationship between literary erudition, empathy, dictatorship and violence? Interestingly Mussolini was a writer too (he published a novel), and Mao worked (like me) a
55.
by YorkshireSeason 5y ago
https://github.com/riscv/riscv-code-size-reduction is the place to check.
56.
by YorkshireSeason 5y ago
> ILP to order programs by generality Sorry, naive question: does ILP test candidate programs by increasing or decreasing generality?
57.
by YorkshireSeason 5y ago
"To see what is in front of one's nose needs a constant struggle. One thing that helps toward it is to keep a diary, or, at any rate, to keep some kind of record of one's opinions about important events. Otherwise, when some
58.
by YorkshireSeason 5y ago
It should be emphasised that inductive programming is not tied to logic programming, and works for every other programming paradigm as well, e.g. functional programming [1, 2]. We could also do IP for imperative programming, although, as
59.
by YorkshireSeason 5y ago
> doubts about the usability of any specification language that ... You can extrude hidden state in ML-like languages, which gives rise to all manner of subtle behaviour in combination with higher-order functions. As Pitts, Odersky, St
60.
by YorkshireSeason 5y ago
If you are interested in Harper's opinions of dynamically typed languages, I recommend his blogpost "Dynamic Languages are Static Languages" [1]. It was also discussed on HN [2], in the context of a blogpost of Tratt'

More ›