The ACM should probably have run this by a statistician before publishing it. As another commenter pointed out it opens with bad experimental design, then goes on about random allocation in a way that makes it clear that the author doesn't understand the subject well.
You don't need sequences of independent samples for assignment. You need sequences that are independent of the sequence of subjects being assigned. If you have a very small number of possible sequences then you can't get independence by the pigeon hole principle, but both 2^(64*N) (from N random 64 bit numbers) or 2^64 - 1 (for N samples from a PRNG) are so large that it doesn't matter.
The argument about a sample of seeds not producing a flat histogram is also weird. Of course it doesn't. That's irrelevant to whether uniformly sampling a sequence of N 64 bit numbers from the space the PRNG provides produces independent sequences from the subjects.
You're describing a very small number of companies that all copied each other's systems. The idea of a terminal role, for example, is pure Facebook. These do not apply in general across the industry except where managers from those small number of companies came in and shoved them in before they were fired.
This seems overly pessimistic. We regularly engage with other cultures and their texts and understand them, and, yes, it takes time and knowledge of context to do so. Someone needs to explain quite a bit about Roman society under Augustus for you to understand what is going on in detail in Ovid's Amores, and the Epic of Gilgamesh is pretty bizarre unless you know quite a bit about ancient Mesopotamia. The Tao te Ching suffers from being written in a style that was very much for people in the know in a certain milieu, limited texts, and a huge amount of cultural baggage on top. The most interesting recent scholarly translation I know of is by Victor Mair, of a different, recently discovered text, and his contention is that the book is a 'mirror for princes' and not a mystical text at all.
> they are still fighting over what 'Hwaet' means
I don't think anyone is particularly fighting over what it means, just how to translate it when there isn't a parallel in modern English. My personal favorite is a translation that opens with 'Bro!'.
> Copy-paste his post into any LLM and ask it whether the post is contradictory or whether it's ambiguous whether this is production-grade software or not. No objective reader of this would come to the conclusion that it's ambiguous or misleading.
That's hilarious! You might want to add a bit more transition for the joke before the other points above, though.
Indemnification only means something if the indemnifying party exists and is solvent. If copyright claims on training data got traction, it would be neither, so it doesn't matter if they provide this or not. They probably won't exist as a solvent entity in a couple years anyway, so even the question of whether the indemnification means anything will go away.
I’m not sure what you mean by a
change of basis making a nonlinear system linear. A linear system is one where solutions add as elements of a vector space. That’s true no matter what basis you express it in.
All these transforms are switching to an eigenbasis of some differential operator (that usually corresponds to a differential equation of interest). Spherical harmonics, Bessel and Henkel functions, which are the radial versions of sines/cosines and complex exponential, respectively, and on and on.
The next big jumps were to collections of functions not parameterized by subsets of R^n. Wavelets use a tree shapes parameter space.
There’s a whole, interesting area of overcomplete basis sets that I have been meaning to look into where you give up your basis functions being orthogonal and all those nice properties in exchange for having multiple options for adapting better to different signal characteristics.
I don’t think these transforms are going to be relevant to understanding neural nets, though. They are, by their nature, doing something with nonlinear structures in high dimensions which are not smoothly extended across their domain, which is the opposite problem all our current approaches to functional analysis deal with.
"Time Enough for Love" is the only Heinlein I've felt any inclination to reread in the past few decades that held up at all (and it's still a good read).
And many people in ads are certainly not working that schedule, and discourage those around them from doing so since it slows everything down long term.
"Zero passion 10am-4pm" doesn't necessarily describe not working hard. You can only do three to five hours of sustainable deep work a day unless you're a freak of nature. Usually anyone saying they do more just pads it out with breaks on Hacker News or Facebook, or has failed to automate the repetitive work in their job. A focused, organized engineer can pretty well exhaust their sustainable capacity plus do their overhead in 10am-4pm. As for zero passion, why do you need passion? And what does it have to do with not slacking?
Facebook makes you demonstrate your impact to the company twice a year. You would be hard pressed to be a slacker long term there.
The work does appear different because at that scale you spend a lot of time tracing out what's there, figuring out context, and trying to figure out how to make a change without bringing down everything else. If you're used to a small company or a small codebase where you spend most of your time writing new code it seems less productive, but it's mostly just different.
You don't need sequences of independent samples for assignment. You need sequences that are independent of the sequence of subjects being assigned. If you have a very small number of possible sequences then you can't get independence by the pigeon hole principle, but both 2^(64*N) (from N random 64 bit numbers) or 2^64 - 1 (for N samples from a PRNG) are so large that it doesn't matter.
The argument about a sample of seeds not producing a flat histogram is also weird. Of course it doesn't. That's irrelevant to whether uniformly sampling a sequence of N 64 bit numbers from the space the PRNG provides produces independent sequences from the subjects.