Kenny

Karma: 648

Kenny 19 Apr 2021 21:04 UTC
1 point
on: What are some good resources on workflow or productivity for mathematicians?
What kinds of advice or ideas do you think would be helpful that’s specific to mathematicians? Any (hypothetical) examples? Are there things that are currently ‘painful’ or that you expect to be, that you don’t think your (hypothetical) workflow/productivity system addresses?

Possibly helpful: I use GitLab for all of my projects; not just software. I find GitLab issues work very well for all but the biggest/longest projects (and even then it’s easy enough to split up a project into ‘sub-projects’ with separate GitLab issues). One reason I like GitLab is that it has a very nice Markdown dialect and it includes pretty good (for me) ‘math’ support. (My favorite part of it’s Markdown is todo lists, i.e. checklists. I find that very useful for, first, outlining what I want to do, and then, later, recording that I’ve done all of those things.)

Kenny 13 Apr 2021 16:34 UTC
1 point
in reply to: FireStormOOO’s comment on: Strong Evidence is Common
This comment is insightful!

The areas where strong evidence is common are largely those areas we don’t intuitively think of as governed by probability theory and where classic logic performs well.

I’m pretty sure that statistics (as mathematics) all assume ‘logic’ (first-order logic at least), so I think this is also technically correct!

Gathering enough evidence sometimes allows reasoning to be performed using propositional logic with acceptable results.

Yes! Being able to use logic can be a fantastic super-power (when it works). Sometimes the universe really is like a Sudoku puzzle!

Being able to use both probabilities and logical statements, and appropriately, is a significant part of what I think David Chapman is gesturing at with what he calls ‘meta-rationality’. And beyond both of those formal rational systems, there’s an entire Platonic universe of alternative ontologies that can also be useful in some contexts (and for some purposes).

Kenny 13 Apr 2021 16:24 UTC
1 point
in reply to: Linch’s comment on: Strong Evidence is Common
This a really nice ‘sharpening’ of ‘fishy’!

Kenny 13 Apr 2021 16:20 UTC
1 point
in reply to: Razied’s comment on: Strong Evidence is Common
Well, yes, if you interpret a lot of thought experiments literally, the proper response is more like “I think I’m having a stroke or that I overdosed on potent psychoactive substances or am asphyxiating.” than anything ‘in the spirit’ of the experiments.

But it gets old fast to describe (yet again) how you’d answer any question posed to you by Omega with “You’re a figment of my imagination.” or whatever.

Kenny 7 Apr 2021 16:03 UTC
1 point
in reply to: Kronopath’s comment on: Your Cheerful Price
I’m glad you thought so!

Your criticism is very fair too. And I’m generally curious about why people ‘bounce off’ the “rationalist community”. I’m also mostly a lurker, particularly IRL. And I think a big part of that is the kind of thing you described. But I do want to do better at being open to really trying weird ideas (and in real life too!). (I’m pretty weird to my acquaintances, friends, and family already.)

I’ve already found this ‘trick’ pretty useful. I haven’t had anyone offer a (radically) honest answer to my asking them for a cheerful price. I suspect that the people I’ve asked don’t fully understand that the question is sincere and shouldn’t be answered in the context of ‘standard’ social norms. And that’s too bad! I’ve asked because I’m serious and sincere about wanting to remove any obstacles (or as many as possible) to us making a particular exchange.

Kenny 18 Mar 2021 0:57 UTC
1 point
in reply to: AnnaSalamon’s comment on: Covid: CDC Issues New Guidance on Opening Schools
From my personal perspective, there is absolutely no Slack in the schools; they’re struggling to do whatever it is they’re trying to do now.

Asking anyone to retrofit all the HVAC systems in schools just … doesn’t seem like something that’s possible for these organizations to do.

Kenny 18 Mar 2021 0:54 UTC
1 point
in reply to: gwern’s comment on: What’s a good machine learning course for _testing_ basic machine learning code?
Thanks! Of course you would know :)

Kenny 17 Mar 2021 22:26 UTC
1 point
in reply to: Kenny’s comment on: What’s a good machine learning course for _testing_ basic machine learning code?
The whole paper was great – I’m sold on property testing!

Kenny 17 Mar 2021 22:18 UTC
3 points
in reply to: philh’s comment on: Direct effects matter!

Similarly, “money makes people better off” isn’t a direct effect of giving them money—it’s mediated by the fact that people can use that money to buy things.

And even that’s not quite completely true. I’m pretty sure people are generally happier immediately after you give them money, even before they spend it! (They correctly anticipate being able to use that money as you describe, i.e. buying or paying for things.)

But I also worry that “this is the immediate effect” is trying to justify them on grounds that don’t work?

Yes, that’s my ‘concern’ as well. Even when ‘direct’ effects aren’t explicitly stated, it doesn’t seem like that’s always, or even mostly, because those effects are being ignored as much as that (implicitly) they shouldn’t be the dominant consideration. That’s a harder thing to disentangle generally.

Kenny 17 Mar 2021 22:12 UTC
3 points
on: What’s a good machine learning course for _testing_ basic machine learning code?
Someone suggested these resources:
- UCI Machine Learning Repository
- MLPerf, which is apparently now part of MLCommons
- Stanford DAWN Deep Learning Benchmark (DAWNBench) ·
- deepmind/lab: A customisable 3D platform for agent-based AI research
They specifically suggested “ALE/DMLab for DRL” but I couldn’t find an obvious result for “ALE” when I searched for it. (It’s a common acronym!)

Kenny 17 Mar 2021 22:10 UTC
1 point
in reply to: Zac Hatfield Dodds’s comment on: What’s a good machine learning course for _testing_ basic machine learning code?
Your paper is excellent so far – very readable! Thanks again!

Kenny 16 Mar 2021 21:42 UTC
2 points
in reply to: cata’s comment on: Direct effects matter!
Yes, that’s a good point – and it’s probably, at least partly, a ‘selection effect’ too, i.e. parents that actively seek advice are probably those that are (overly) worried about medium and long term effects/consequences. And it’s certainly the case that some parents, or all parents some of the time, neglect important the medium and long term.

But there is parenting advice along the lines of what I endorse – and it’s great! And I think many people do support a lot of things along those lines, if only tacitly, e.g. ‘free range’ kids kinds of things.

(And of course I’m incredibly sympathetic to parents that are terrified of harming their children in any way, especially inadvertently. It really is an awesome responsibility!)

Kenny 16 Mar 2021 21:36 UTC
1 point
in reply to: habryka’s comment on: Direct effects matter!
Thanks for the clarification! I think your “mod note” comment serves as an effective and eminently sensible ‘warning’ anyways.

Kenny 15 Mar 2021 22:32 UTC
1 point
in reply to: habryka’s comment on: Direct effects matter!
I’m not sure what you mean by “this is comment is fine”. I don’t think it’s a good comment, but I don’t have strong feelings about it, e.g. being deleted.

Kenny 15 Mar 2021 22:31 UTC
8 points
in reply to: AllAmericanBreakfast’s comment on: Direct effects matter!
There are some pragmatic considerations to this, e.g. ‘harshing their high’. I wouldn’t think people that drink alcohol are unaware of the health risks (and other complications), so bringing it up would pretty reasonably be interpreted as criticism, or your trying to convince or persuade them to give up something they enjoy and choose to do despite the risks.

But I’ve had interesting discussions with stoners about the possible psychiatric issues of marijuana, and drinkers about their health issues due to drinking, so I’m not sure that it’s ‘unacceptable’ – at all anyways. I can certainly appreciate why anyone wouldn’t want to be ‘badgered’ about the risks of something they still choose to do! (I find it annoying when people do it to me.)

I also wouldn’t think it unreasonable that, ex. parents watching their children play American football wouldn’t also want to, at that time, talk about the attendant risks of head injuries. At least in that case, I’d presume they were already aware of the claims about those risks and had made a decision to let their kid(s) play anyways.

Kenny 15 Mar 2021 22:23 UTC
6 points
in reply to: cata’s comment on: Direct effects matter!
I very much, personally, disagree with the parenting philosophy you’ve outlined! (I’m NOT also claiming that you share that philosophy. I hope you don’t!)

Children – and everyone is one! – only have one life! There is no ‘long-term’ – when does that start? At 30? 40? 50?

The long-term includes now! It started in the past actually. And the point really is (or should be) to ‘maximize the integral’ of ‘living a good life’, not preparing for some ‘future’ that never arrives.

As a parent myself, I want my child to live a great life, including today, tomorrow, both short-term, and long. I don’t want to reward them only because of a practical “you can only push so hard” limitation on the amount of unpleasantness they can bear.

I also want to push back on this being “100% clear … in the eyes of society”. There are others that disagree with this as I do. And I’d expect most people do too, to at least some extent. People’s ‘revealed preferences’ certainly don’t seem to match these supposed prescriptions – not generally and definitely not “100%”.

Kenny 15 Mar 2021 22:15 UTC
4 points
on: Direct effects matter!
I’ve been thinking about this – thanks!

One problem is that ‘direct’ or ‘first-order’ is ultimately nebulous. There is no unique direct effect or first-order consequence to anything, i.e. it depends crucially on context, scale, scope, etc..

Consider the direct effects or first-order consequences of a heroin overdose. Restricting ourselves to the macroscopic, i.e. human-scale, I’d imagine that a heroin overdose is initially very positive! It does seem a little … insensitive? besides the point? to emphasize “Heroin overdoses are really great until you start to die shortly thereafter!”.

A part of not emphasizing the direct effects or first-order consequences is an implicit claim that, whatever they are, they are swamped (or maybe simply just outweighed) by subsequent effects or consequences.

Hence the long-term focus of a lot of your examples, e.g. unemployment insurance, basic income.

Similarly, consider rent control – it really is awesome and amazing! At least it is for the people that ‘win the lottery’ and get to live in a rent controlled unit. And yet, overall, it’s a terrible deal for everyone.

But I’m not against including all of the effects and consequences, including the ‘direct’ or ‘first-order’ ones – we should shut up and multiply (or add, or whatever).

Kenny 14 Mar 2021 19:29 UTC
1 point
on: What’s a good machine learning course for _testing_ basic machine learning code?
I started working on re-implementing my Octave code for the course I took previously and it’s going pretty well!

Kenny 14 Mar 2021 19:26 UTC
1 point
in reply to: Zac Hatfield Dodds’s comment on: What’s a good machine learning course for _testing_ basic machine learning code?
I see now that my question title could be better. I’m more looking for test cases, than testing tools.

I’ve added your paper to my reading queue! Thanks!

Kenny 14 Mar 2021 18:32 UTC
1 point
in reply to: bouilhet’s comment on: Magical Categories
Yes; good points! Do note that my original comment was made eight years ago! (At least – it was probably migrated from Overcoming Bias if this post is as early as it seems to be.)

So I have had some time to think along these lines a little more :)

But I don’t think intelligence itself can lead one to conclude as you have:

If it is intelligent, it will make the distinction.

It’s not obvious to me now that any particular distinction will be made by any particular intelligence. There’s maybe not literally infinite, but still a VAST number of possible ontologies with which to make distinctions. The general class of ‘intelligent systems’ is almost certainly WAY more alien than we can reasonably imagine. I don’t assume that even a ‘super-intelligence’ would definitely ever “differentiate between smiley-faces and happy people”.

But I don’t remember this post that well, and I was going to re-read before I remembered that I didn’t even know what I was originally replying to (as it didn’t seem to be the post itself), and re-constructing the entire context to write a better reply which my temporal margin “is too narrow to contain” at the moment.

But I think I still disagree with whatever Shane wrote!