emsal

Doesn't this fundamentally misunderstand the Pareto principle? The 80% and 20% of causes in standard examples don't refer to portions of a sequential effort, but rather slices of competing agents/producers/customers in an economic system. Like, 20% of clients account for 80% of sales, that kind of thing.

I think the "repulsive attitude" I'm referring to is shouting "Just grow a thick skin!" to someone who is saying that they're suffering from abuse. In some cases, the suffering is so acute and so deep ("bullets") that it's not necessarily actionable on the part of the person suffering; rather, it's a call for help that others may respond to.

OP replied to my comment and I now see that it makes sense as an observation. The narrative I was referring to is something that I see common in discussions surrounding abuse; "there's nothing to be done about it; [people should] just grow a thick skin." If we agree that this exists, then I posit that it's a toxic thing to say to someone who's coming forward and saying that they're suffering from this. It's also toxic to use that as a way to dismiss measures that we can build in technologically to combat abuse, like building blocking mechanisms or modifying them to make them better. Hope that clarifies things.

Perhaps I replied to the wrong comment then. Sorry for being confrontational.

Unfortunately HN is no longer a place where you can assume good faith observations when something like that is said. OP's words play into toxic narratives that need to be called out for the sake of other people reading the comments. I might be wrong about OP's intentions when giving my reply, but I am pushing back against a narrative, not against a specific person.

You might be correct but it still stands that nothing is added to the discourse by just telling people to grow a thick skin. It's a repulsive attitude and I had to call that out.

"hey reminder that" does make a difference, but that difference is invisible because it's exclusively in the form of incidents that did not happen. Remember that everyone reading the comment is a human too and some will be receptive. Good is good.

Many public figures have social media and PR teams to help deal with stuff like this, you really expect that every individual out there should just grow a thick skin as culture is continually encouraging us to have more and more of a public presence? This is the kind of thinking that gives a free pass to entrenched patterns of discrimination and targeted hate speech. It's very possible that emails like this wouldn't just be a spew of random insults but targeted towards a person's very specific triggers, and in that case a thick skin cannot stop a bullet.

Perhaps this is the obvious comment, but I really hope that something that employs technology like this can get off the ground and become a services company. The ideas about democratizing the AI inference hardware space and making it energy efficient really resonate with me.

https://youtu.be/hph9OeKjg3w -- here's a video about this about a research effort to accomplish precisely this.

I once attempted to make something like ProxySwitchy for DNS[1], but I didn't work on it long enough to get off the ground. This article made me think about it again. Is there actually a use case for that kind of thing?

[1]: https://github.com/emsal0/resolvplox

This is awesome! I really wonder if GPs will see an uptick in usage in the future, I've seen a lot of interest in them lately.

Also, thanks for including the source so they can be seen in the browser dev tools, don't know if that gets done enough with interactive visualizations like this.

As imperfect measures as they are, statistical significance measures like the p-value are more epistemologically sound than any arbitrary rule of thumb-threshold on the size of the effect.

> The fragment synthesis procedure uses a powerful, but computationally expensive, algebraic rewrite-based search algorithm to identify non-trivial compositions of blocks. The performance of the algorithm depends on the complexity of the target equation rather than the size of the overall system.

Could the construction of these blocks, as well as the composition step, benefit from training an RL agent on the state space of circuit configurations/the action space of connecting two blocks (or other operations)? Does any facet of the problem make the idea of using reinforcement learning intractable/otherwise a bad idea?

Ideally you'd be able to accurately gauge your progress without having intense negative emotions attached to it. Certainly it shouldn't undercut a person's feeling of validity in the world.

I think "feeling behind" means different things for different people, depending on the experiences in their formative years they have associated with the thought of being behind. The ideal response to the feeling depends on what the feeling is.

That's true, I don't even remember the ending to Snow Crash which I read some ~6 years ago. The unraveling premise that comes about in the ~3rd quarter of the book is what stuck with me. Think it'll be the same for Cryptonomicon.

Tangent, but Cryptonomicon by Neal Stephenson[1] is a good novel that touches on WWII cryptography topics really well, going decently mathematically deep into the mechanics of how they work. I'm about 75% of the way through.

[1]: https://en.wikipedia.org/wiki/Cryptonomicon

For what it's worth, Dr. Robert Sapolsky, who wrote one of the books that this article draws on, has a fascinating set of lectures, free online, from his Human Behavioral Biology course at Stanford[1].

I personally haven't listened to it in very great detail as if I were taking the course, but I often skip around to listen to the dissections of various aspects of human behavior.

[1]: https://www.youtube.com/playlist?list=PLD7E21BF91F3F9683

Key caveat here is that it only works on AMD cards currently.

I'm definitely not as savvy - can you expound on how the decisions that Ballmer and Muglia made paved the way for Nadella's success?

This kind of writing makes me upset.

* It's really self-important. Not only is it selling you on a particular strategy for attaining success, it also tries to sell success in things like public speaking and doing architectural work as an absolutely important part of one's life, and that implicitly a person's existence is invalidated if they aren't constantly trying to achieve this kind of success. It doesn't do it explicitly but the very notion of "real" and "fake" and other words like "wasted" complete with the trite diagrams showing that "hey, all your efforts are going into this small circle" give a very strong implicit value-judgement of the reader.

* There's no proof. I don't know if I'm on the mark with this one, but I think that the act of omitting any sort of data about measuring the outcome of success when taking different approaches seems to imply to the reader that the argument should just "make sense" i.e. it's a truth that the reader already knows, they should just find it within their own observations in order to understand it. Here, have a handful of anecdotes to top it all off in case you weren't convinced. Overall this just feels like it's made to make the reader feel a certain way (motivated) rather than actually teach them any solid information.

* What even is real and fake? The readers are given a bunch of examples and then we're left on our own to figure out what falls into which category. Someone commented on the article saying that if someone wanted to watch and understand anime in Japanese, they could just do that and that'd be the real thing, with the fake thing being taking the time to learn Japanese. This is obviously not going to be successful, so at this point the author's prescription has failed as a framework for achieving success.

---

This kind of fiery motivational content could be harmful as much as it is useful. It'd be fine if an article, devoid of substance as it may be, was only meant to make readers feel motivated, but the problem is that this kind of fiery motivational content does different things for different readers. A person in a bad, self-loathing emotional state could be rendered feeling even worse, thinking that everything that they're doing at present is fake while everything that their peers are doing are more real, even when that's blatantly untrue. The devil's in the details and personally, I'm not going to let myself get affected by this personal philosophy if the case for it is this weak.

What's scary about Googling for specific ML questions beginning to look like googling some Windows problems? This kind of dichotomy already exists between people who interact with ML only through keras and those who use a more flexible library. If you fall into the latter category it's still easy enough to find the answers to ML questions that mostly pertain to the deeper stuff. Don't even get me started on the difference between people who do and don't know statistics theory. But the point is that all of these groups do ML and they coexist just fine.

On another note. I don't think technical proficiency (or whatever your elitism-metric is) is very correlated with using Windows as the primary platform. You can Google stuff related to a Windows problem and through that learn how to work with the registry, and yes, some of the things you need to do require you to RTFM. So this elitism also suffers from being off the mark.

This is really concise, thanks! Removes a lot of the mysticism that is all too present in mathematics writing for laypeople (this explanation is just as understandable).

Theoretical physics is sporadically coming back into my mind again after a lot of time learning CS and Stats.

I'm wondering now, is the goal of theoretical physics sort of like the goal of topological data analysis[0]? Reading about things like how Max Planck came up with quantum theory to explain the measurements of black body radiation reminded me a lot of the ML task trying to fit a model to data.

My bias is definitely showing here but is this a useful framework to be able to think about these things?

[0]: https://en.wikipedia.org/wiki/Topological_data_analysis

I don't think the password managers do, for master passwords. So if you ever consider it, you can set your master password to a strong one generated here and let the pw manager generate those 30 char random strings that fill all of the requirements for the other sites.

There was this listing by a professor of Physics of all the topics you need to study in order in order to become a "GOOD Theoretical Physicist"[1]

[1]: https://www.goodtheorist.science/

I think there's room for both self-directed aggregation (à la reader sites) as well as the centralized aggregator model. It's quite shackling to never have any control over the content that you're about to consume, and being able to check up on websites that you have personally discerned as being good content is quite empowering.

I'm no expert but I think that the async-await pattern is quite similar to CPS.

If, in addition to the Morse Code associations with letters, you have a dictionary of permitted words, this becomes a fairly straightforward DP, does it not?

The one I learned programming from first is A Byte of Python by Swaroop CH: https://python.swaroopch.com/

Really concise introduction to programming in Python; easy enough for someone to absolutely devour and learn incredibly quickly.