All of mraxilus's Comments + Replies

The 30 was crossed out in the original quote. (between 30 15 and 120 minutes of clean dialogue for each character). I guess quoting it didn't take the formatting with it.

No problem, let me know which ones you find the most interesting. I'll try to improve the quality per link over time.

Is that something an individual user can set up for themselves or do we have to message an admin for that?

2habryka
You currently have to message an admin for that. Mostly because we are too worried about people spamming things if we allow them to do this automatically.

Awesome, thanks! Is there an e-book version?

Great post, but I would suggest avoiding cliche titles such as 'The Unreasonable Effectiveness of X'.

It provides a lot less information about what the post contains than a more carefully crafted title would, for not much gain (maybe people are more likely to read TUEoX posts?).

Basically, I always feel a discomfort when people have 'TUEoX' as a title, and never provide a strong argument for why the effectiveness of X is actually unreasonable (i.e. beyond the limits of acceptability and fairness).

Sure, it's effective yes, and it certainly was unexpected, but is it fair to say it's unreasonable?

1Richard_Ngo
I confess that I have a weakness for slightly fanciful titles. In my defence, though, I do actually think that "unreasonable" is a reasonable way of describing the success of neural networks. The argument in the original paper was something like "it could have been the case that math just wasn't that helpful in describing the universe, but actually it works really well on most things we try it on, and we don't have any principled explanation for why that is so". Similarly, it could have been the case that feedforward neural networks just weren't very good at learning useful functions, but actually they work really well on most things we try them on, and we don't have any principled explanation for why that is so.
We know that students can’t move on to algebra until they have perfectly memorized their times tables.

That's how most people progress into algebra, but I don't think you need perfect memorization of the times tables, simply the ability to reproduce it.

As a fellow programmer, I think the epsilon fallacy is more memorable. If it were Amdahl's fallacy, it would be one of those fallacies I have to constantly lookup the fifty times or so (terrible memory, and not enough slack/motivation for a fallacy memory palace).

There would come a point where, if a large enough amount of people believe it, it would start affecting you. Would make it harder to find jobs, to find other people to discuss ideas with, to convice people of an argument that relies on statistical significance, etc. It would have a huge effect on economical progress if the majority of people started believing that math is of no use to anyone.

I initially read it in the same way you did, however I also think SquirrelInHell has a point. But I would say the place where he's going wrong is sometimes (possibly most of the time) people don't know what it is they're seeking from a conversation. A lot of people don't know themselves well enough, so having been promted with that question allows them to properly introspect, perhaps.. but I do agree that password based conversations are frustrating.

I'd like to know more about the argument for why tribalism should be indulged rather than surppressed. Anyone got any good links for the topic?

3Viliam
I know that in psychology almost nothing replicates, but it seems likely to me that most attempts at suppressing emotions lead to either their occassional outbursts, or mere denial. An archetypal example would be a person screaming with red face about how they are not angry -- that seems obviously harmful, even more so for a wannabe rationalist. On the other hand, if you redirect your anger into e.g. exercising, you can improve your health. If people could reliably turn off their tribalism by pressing a button... I guess I would still have some doubts about turning off something that has an obvious evolutionary purpose... but I would be okay with other people doing potentially harmful experiments on themselves. But seems to me that when people e.g. avoid expressing membership in the rationalist tribe (sometimes proclaiming that they want to keep their identity small), what usually happens is that they already identify as members of some other tribe, and understand that double membership would create a few conflicts. It would be more honest to admit that, though. But if you know that your brain is designed to create strong feelings about this or that tribe, and that those feelings can e.g. motivate you to overcome akrasia and do stuff that is approved by given tribe, you might harness that power by choosing a tribe whose goals appeal to you. Please understand that I am not saying that you should maximize the feelings of tribalism, or anything like that. Instead, simply accept that they exist, and choose a useful channel for them to flow in, as opposed to letting them find their own channels.

I for one, would be all for a swap over to a no caps English, but my desire to not cause such a ruckuss as this over trivial aesthetic things out weighs it. If you think of all the extra effort/time you waste breaking your typing flow by having to capitilise letters at the beginning of sentences. Isn't that what the period is supposed to before, to denote the end of a sentence and the start of a new. I do think capitals should be used for emphasis though, and for proper nouns if they intentionally choose to have a capital letter. Such as LaTeX.

I actually used control-f for every letter in the alphabet before I read the last sentence... Touché. Reminds me of an old exercise in school where there are a bunch of random instructions on a handout and the last is to ignore all previous instructions. Moral of the story was to read all questions before answering them. A more useful one would be to don't get to focused on what's right in front of you, explore a little bit first to ensure that you're taking the best immediate action.

4alkjash
I remember exactly the same exercise from elementary school, because I was the last one to catch on.

Possibly because at first glance your writing comes across as remarkably similar to poetic profundities without substance. Perhaps, try being more straightforward with your definitions, assertions, and claims.

0Zane Scheepers
I made no claims. I simply had an idea I wished to share. I've since discovered that someone else already had this idea. In debates, you don't get points if someone already agrees with you prior to the debate. You get points if someone changes your opinions. Allocating points to common beliefs simply reinforces the belief, true or not. If 99 out of a hundred people agree that we can't see light, they would never delve deeper into why we can't.

I would agree, a lot of DL concepts became much clearer after I worked with tensors enough to be able to manipulate them in my mind. Before that I just sort of just made sure the shapes matched up and hoped for the best. Also, you tend to see a lot of beginner questions/confusion on DL libraries (Keras especially) about tensor shape mismatches, suggesting that it is quite a common problem area.

Fair point, I never really considered that. It also tends to be the path followed by a large portion of our population.