avancil

Karma: 112

avancil 23 Mar 2024 0:10 UTC
0 points
−1
on: [Linkpost] Vague Verbiage in Forecasting
Ironic that “Maybe” seems to have one of the narrower ranges of probabilities...

avancil 7 Mar 2024 21:53 UTC
34 points
28
in reply to: Ben Pace’s comment on: Using axis lines for good or evil
The cell borders example is misleading. The readability issue is not the cell borders themselves, the issue is that the borders are heavier weight than the text, and there’s no difference between the borders separating the row and column headers, and the borders separating the data rows and columns.

If your only choices for gridlines are “off” and “obnoxious”, “off” seems like a good choice. And for small tables, no borders works well. But for larger tables, finer lines (maybe in a lighter color or shade) can really improve the readability.

avancil 25 Feb 2024 7:53 UTC
5 points
0
on: Balancing Games
In multiplayer games, one balancing factor is that other players can gang up on the person who is ahead. Depending on the game dynamic, this can even things out a lot. In some games, this even creates the dynamic where you don’t want to look too strong, so that others don’t focus their attention on you.

Playing games against my kids when they were young, rather than just slack off and let them win, it was more fun for me to figure out the best way to handicap myself: What algorithm for sub-optimal play would keep the game close? Solving that puzzle effectively became my victory condition, rather than the game’s victory condition, and I was effectively competing against myself, a more balanced opponent.

avancil 24 Feb 2024 20:40 UTC
7 points
2
in reply to: jasoncrawford’s comment on: Why you, personally, should want a larger human population
The question of what IS happening versus what SHOULD happen with population growth are certainly two different things. My point is that arguments for growth ultimately need to address the questions of how big should we grow, and what happens when we reach that point. If our economy depends on continued growth, that’s going to stop working at some point.

While the physical limits of the universe are a long ways off, there are other limits that we could hit much sooner. Underlying your pro-growth arguments, there is an assumption that collective intelligence can continue to grow without limits, leading to technology that can grow without limits. I would question those assumptions.

And of course, your post is ignoring the costs of growth. Ideas are non-rival goods, but space on this planet, and physical resources, are rival goods. If intelligence (and the resulting technology) reaches a point of diminishing returns, but the costs of growth hit an upward inflection, you quickly hit a limit. For example, larger, more complex systems risk becoming less stable, while coordination problems can grow factorially.

Reasonable people can disagree on whether the current population is too big, too small, or about right, but “ever larger” is not going to work as an answer. At some point, we need to either figure out how to have a stable population, or deal with the less pleasant alternatives.

avancil 23 Feb 2024 22:56 UTC
15 points
17
on: Why you, personally, should want a larger human population
The fundamental problem with these anti-Malthusian arguments is they ignore the fundamental reality that exponential growth is unsustainable. There are physical limits to the universe, whether you believe the earth can support 10 billion or 100 trillion, or if we can expand through the universe and achieve a billion billion times more than that, it won’t take that long, with exponential growth, to get there. At a certain point, the entire mass of the universe has been converted to human flesh. Some point before that, we either stabilize, or collapse.

Maybe we get past the obvious physical limits by ditching human bodies and uploading to the matrix. But, that only gives us a few more orders of magnitude.

And note that sub-exponential growth is mathematically equivalent to saying that the growth rate approaches zero.

So, it becomes a question of what the ideal max population level should be. The answer will clearly change based on available technology, but is not unlimited.

avancil 12 Jan 2024 5:49 UTC
1 point
−2
on: Introduce a Speed Maximum
Where I live, I don’t see many people going 15 over. I see most people going within −3 to +9 of the speed limit. They’re following the law—maybe not strictly as written, but as socially understood. There are a few people breaking the law—and they get ticketed, etc. There are places where the speed limit is unreasonably low, and gets ignored (e.g. speed limit drops from 75 to 55 for construction zone, but no construction activity is visible), but in general, people around here follow the speed limit—as socially understood.

The social definition is that +9 is OK, and it is logically based. Speed limits are only given to the nearest 5 MPH, so arguably (not legally arguably, but in people’s heads) going 4 MPH over is not really speeding. Then, you need some kind of reasonable cushion. It defies common sense (but not the law) that a certain speed is perfectly legal in a variety of reasonable conditions, but 1 MPH faster than that is a crime. So, you have to have reasonable padding, say one 5 MPH increment. Thus, if you’re going within 9 MPH of the speed limit, you’re not speeding, just bending the speed limit, pushing into that safety margin.

Bumping a speed limit from the current value of 55 (+ social tolerance) to 65 (no tolerance) would not work. It would quickly become 65 (+ social tolerance), because you have not addressed the arbitrary nature of the speed limit. With strict enforcement, you could reduce the social tolerance value (I believe it was smaller before the 1970s national speed limit of 55 fostered widespread disrespect for speed limits), but it would difficult to eliminate it.

avancil 12 Dec 2023 5:56 UTC
3 points
2
in reply to: faul_sname’s comment on: Deeply Cover Car Crashes?
Yes, when it comes to ordinary driving situations, there’s only so good you can get, if you can get from A to B without trouble, without annoying and/or scaring your passengers or other people on the road, it’s hard to do noticeably better. It’s hard to get too much above the median; the 80th percentile driver won’t seem that different from the 50th percentile driver. But, you can be really bad and drag the average down. Thus, the average is below the median, ergo most people (most drivers, anyway) are above average drivers. (Even assuming we are using some identical, objective scale, which, as jefftk points out, is not going to be the case.)

avancil 24 Nov 2023 19:18 UTC
2 points
−1
on: What did you change your mind about in the last year?
At the beginning of the year, I had never heard of ChatGPT, and thought AI would continue to progress slowly, in a non-disruptive fashion. At this point, I believe 2023 will be at least as significant as 2007 (iPhone) in terms of marking the beginning of a technological transformation.

avancil 25 Jul 2023 5:44 UTC
3 points
0
on: Secure Hand Holding
Just be careful your secure grip doesn’t inadvertently cause the child an arm injury. “Nursemaid’s elbow” and certain types of radial fractures can result from a hard yank on a child’s arm.

avancil 26 Jun 2023 16:58 UTC
7 points
0
on: Model, Care, Execution
Unendorsed while acting/Endorsed reflectively: This is not so strange a failure mode as you are indicating. You take some action, which at the time you know isn’t right (execution failure). Later, you come up with some post-hoc justification for your actions. This is another failure mode to be aware of, affecting the postmortem.

Postmortems can have their own failure modes. We might disagree about the facts of what happened. We might have different perspectives on the context in which things happened. We might struggle to be honest with ourselves about our motivations at the time, or our analysis of past events is colored differently by information gained after the events—maybe giving us better insight, or maybe clouding over the original failure.

avancil 26 May 2023 4:51 UTC
1 point
2
on: Mob and Bailey
Internally, though, people are not monolithic. In some ways, they are more like a mob. Even debating with one person, you can run into the same type of difficulty.

avancil 2 May 2023 3:46 UTC
1 point
1
in reply to: Dagon’s comment on: What Boston Can Teach Us About What a Woman Is
Upvoted. While I would hate to see this forum become another battleground for culture wars, I also don’t want to see people avoiding topics because they touch on culture war stuff. Using Rationalist tools to deconstruct the dynamics of these battles (without fighting the battle) seems like a useful thing.

While I don’t necessarily agree with the post, and I can see where it might, unfortunately, trigger some negative responses from some, I found parts illuminating, and it got me thinking about some useful reframing tools. All too often, people in the culture wars talk past each other because they are using the same words to point to different things.

avancil 30 Mar 2023 23:53 UTC
1 point
0
in reply to: dr_s’s comment on: The salt in pasta water fallacy
It’s hard to get good data, yes, particularly in a politically charged environment. But, I would have liked to have seen some evidence that for a given mitigation, our leaders tried to get a best estimate (even if it is not a great estimate) that it will prevent X COVID deaths, at a cost of Y (dollars, QALYs, whatever), and had done some reasoning why utility(X) > cost(Y). We might disagree on the values of X and Y, and how to compare them, but at least we would have a starting point for discussion. Instead, we got either “Don’t take away muh freedom!” or “We must stop anyone from dying of COVID at all costs!”. And a lot of people died, and we inflicted tremendous damage on ourselves, while doing some things that were maybe beneficial, and a lot of things that were clearly stupid, and we’re not in a position to do better next time.

avancil 27 Mar 2023 21:12 UTC
5 points
0
on: The salt in pasta water fallacy
There have been multiple egregious examples of this fallacy with respect to pandemic policy. The complete lack of rational cost benefit analysis (across the political spectrum) for the various measures was truly disheartening.

But to give a less politically charged example: Locally, in response to drought conditions, some restaurants announced they would only bring glasses of water to the table, on request. Now, this might make some sense in terms of reducing labor, although the extra work of having to ask everyone, and possibly bring out additional glasses later, probably cancels this out. But for reducing water usage, this is just silly. Let’s do some math:

Total water usage in US is about 1000 gal/person-day (including domestic, agricultural and industrial uses). So, assuming 50% of patrons leave their 12 oz glass of water untouched, you are reducing the daily water consumption of restaurant patrons by less than 0.005%. Of course, it also takes water to wash said glass—it can vary a lot, but 18 oz of water would be a reasonable estimate (assuming an automatic commercial dishwater). That gets us up around 0.012%, or the water you get from running the faucet in the sink for 5-10 seconds.

avancil 1 Mar 2023 4:30 UTC
37 points
4
on: Contract Fraud
I remember as a kid, about 12, loaning my less-mathematically-astute younger brother $4, at 10% interest per day, compounded daily. I remember gloating about how much money he was going to owe me. I was going to be RICH, mwuahh hah hah!!

My Mom told that loan sharking was illegal, and my Dad told me that contracts with minors were not enforceable. My brother I think borrowed some money from one of his friends (on much more favorable terms), paid me back with one day’s interest, and never borrowed money from me again.

avancil 17 Jan 2023 0:55 UTC
1 point
0
in reply to: Gerald Monroe’s comment on: On Cooking With Gas
I’m not sure how I would use my round-bottom wok on an induction burner, but maybe there’s something that would make it work? And how I would char the skin on a chili pepper?

The converter plate is an interesting idea.

Maybe my ideal stove has three induction burners and one gas. Maybe I would discover that I rarely used the gas burner, and decide all induction works fine, and that I could just use my camp stove or outdoor grill if I actually need a flame. A hybrid stove could be a useful approach to getting people to overcome their hesitancy in switching to electric. Admittedly, it wouldn’t be very practical in most cases, only really feasible in kitchens already built to support both electric and gas. (My kitchen has a gas stove, but actually does have the wiring for an electric stove.)

avancil 13 Jan 2023 18:20 UTC
3 points
2
on: On Cooking With Gas
My ideal stove (which I don’t think anyone makes, for various practical reasons) has three burners, one gas, one induction and one electric.
- Induction because induction is interesting, efficient, clean, and works well with ferrous cookware
- Gas because induction doesn’t work with some cookware, and sometimes you need an actual flame
- Electric because gas burners suck at low temperature operations—they get hot spots—and induction doesn’t work with some cookware.
Three burners because that’s the most I can effectively use at one time. But maybe having four burners (with a second induction or gas burner) would make sense.

avancil 20 Aug 2022 22:31 UTC
2 points
0
on: Vague concepts, family resemblance and cluster properties
Vague concepts are a...vague concept.

avancil 28 Jun 2022 2:38 UTC
3 points
5
in reply to: Adam Zerner’s comment on: Air Conditioner Repair
This strategy works for something that happens once, but for something that could be a pattern (e.g. getting ripped off by contractors), allocating thought to it would be worthwhile—but only if you are focused on learning from the experience, and avoiding this type of problem in the future, as opposed to just wallowing in the fact that you were wronged. (And that’s also easier said than done.)