= 27d567bc2d48d9f40e9ccc20ea370b15)
RichardKennaway comments on Rationality Quotes: July 2010 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (216)
Jonathan Swift (also attributed to Pope)
Abraham Lincoln
=> A man who is not ashamed of being wrong will be thought of much by Abraham Lincoln?
As an aside, one should be ashamed of being wrong when there was sufficient time and information to be right, and still admit being wrong. It should lead to better care taken next time.
You should resolve to take better care next time, and evaluate where your mistake was, but I don't see how shame helps anything.
Getting singed helps avoid hot objects. Shame helps avoid stupidity.
Shame leads to a variant on guessing the teacher's password-- an effort to not piss people off, without asking them what might be problematic. After all, you're supposed to know better than to make that mistake.
"The shameful does not learn" - Talmud
I agree with your point, but I do not think the emotion should be dispensed with altogether, if that's really possible. I think you can separate social embarrassment, which in this context is counterproductive, from deep very personal private embarrassment of having been stupid.
For an even somewhat rational person, pain is far stronger than necessary as a warning sign. As someone generally concerned with my own body's welfare, the mental equivalent of popping up a politely worded dialog box would be sufficient. I find that shame is likewise overkill for solving this problem.
Personally, I don't mind pain being a strong enough warning that it's hard to ignore. I can see the need for that.
I think the problem with pain is that it's like those stupid car alarms. You know the ones that should be programmed to turn off after five minutes because any good they might do will have been done by then if it's to be done at all, but they actually keep going all night?
That's what pain should have: a way of saying, after some appropriate enforced time delay, fine, I've got the message, I'm doing everything I can about the problem, you can stop now.
I can't. I've had problems with pains that are demonstrably unrelated to any threat to bodily integrity and for which there is no known technique that removes it. If pain were limited to real threats, I'd agree, but it's not.
So it's not even an issue of "yeah, I get the message, you can stop reminding me"; often times, there is no message to be given, just suffering.
Have you never been tempted to push ahead doing something you want, and ignored a minor pain? We'd just ignore pop-up boxes if we were in the middle of something we considered important.
Absolutely. But this is not a way in which pain is superior to a pop-up box. If the pop-up box that replaced intense pain had alarms and flashing lights attached, and the one for more minor pains did not, I would pay attention to the alarms and flashing lights.
Speaking in terms of real pop-up boxes, you might be surprised at how easy it is for people to ignore the content of even the most blaring, attention-grabbing error messages.
A typical computer user's reaction to a pop-up box is to immediately click whatever they think will make it go away, because a pop-up box is not a message to be understood but a distraction from what they're actually trying to accomplish. A more obnoxious pop-up box just increases the user's agitation to get rid of it.
As rationalists, we try hard to avoid falling into traps like these (I'm not sure if there's a name for the fallacy of ignoring information because it's annoying, but it's not exactly a high-utility strategy), but part of the way we should do that is to design systems that encourage good habits automatically.
I like Firefox's approach; when it wants you to choose between Yes or No on an important question ("Really install this unsigned plugin?"), it actually disables the buttons on the pop-up for the first 3 seconds. You see the pop-up box, your well-honed killer instinct kicks in and attempts to destroy it by mindlessly clicking on Yes so you can get back to work already... but that doesn't work, you're surprised, and that jolt out of complacency inspires you to actually read the message.
I suspect a "Hey, have you noticed that something has penetrated the skin of your left foot?" warning might benefit from having the same mechanism.
I appear to be a mutant: I always read pop-up boxes.
By all means, please adapt my analogy to something that you would actually pay attention to.
That doesn't seem a lot better - I know what it wants me to do with an unsigned plugin prompt - review the site, verify the download hash, review the plugin source code, look for reviews of the plugin, author, site...
What I actually do is wait 5 seconds, then click "yes do it" as soon as possible. So the utility is ... well intentioned, but still ineffective.
It's true that it's less than perfectly effective, but it serves some purpose: I almost always install plugins from the Mozilla plugin site, where a rating is immediately available, and where a virused plugin would probably get removed very quickly. Under those conditions, I know that I'm fairly safe just installing it anyways.
However, a malicious site could attempt to infect my browser by installing a plugin, which is where the timer comes in handy. It could even attempt to hide the plugin dialog with lots of other useless dialogs ("Really submit this comment?" Yes. "Really really submit this comment?" Yes. "Really really REALLY submit this comment?" Yes. "Install this plugin?" Yes. Oh, hold on, wait! Crap.)
More generally, timed dialogs are helpful because they increase the chance that you notice what it is you're confirming. If you know you're doing something risky and want to do it anyways, so be it... but at least you know what it is you're accepting, and are given a greater opportunity to back out if you are surprised by the level of risk.
The about:config option "security.dialog_enable_delay" allows one to reduce the delay to 0.
It seems pretty clear to me that this was not true in our ancestral environment. It may be the case in our present artificially benign environment however.
That is precisely what I mean; but also note that there are circumstances in the ancestral environment in which pain is entirely useless, such as when one has been mortally wounded. So even in the worst case we can do better than pain, and in the current case I suspect we can do much, much better.
Evolution has the problem of path-dependence, though. Once the "don't do that" / "pay attention to that" mechanism builds up slowly over many generations, it cannot refactor in such a way that it surgically cuts out the internal feeling of pain in precisely those circumstances where, "hey, might as well give up".
It's hard to see what reproductive benefit there would be to reduced suffering when dying either so there is unlikely to be any evolutionary pressure in that direction.
Thanks for the explanation. I like the argument, but still willing to play the devil's advocate: popup box is nice when you're paying attention, but it does not produce learning. Imagine working on a plant where a wrong move can cause a serious injury. Popup boxes will not produce the muscle memory needed to navigate.
I would argue this strongly about pain, but I am not sure how well the analogy transfers to mental errors.
To continue the computer analogy, it would have to be a popup box that steals focus, so that you can't do anything else until you acknowledge it's there.
Does pain aid in creating muscle memory? I hadn't heard that before.
Should it pop up again if pain increases from mild to strong? Should it pop up periodically when pain is extreme?
I'd rather prefer to consciously disable drive to remove pain source, but stay informed of kind and intensity of pain.
Edit: AFAIK insects use that kind of pain processing, they react on pain, but they don't get overwhelmed by it.
Sure. The analogy to computers is not a perfect one, because brains don't function like modern computer operating systems. My objection to pain is not that it is uninformative, it's that it's overwhelmingly unpleasant even when we do not wish it to be.
Yes, it can inflict more harm by forcing you into suboptimal decisions. Shame can be alike too. So, I vote for insect-like shame and pain processing, I've mentioned in grandparent.
Well, it would have to be a popup box that interrupts whatever you're currently thinking about. It grabs focus, to continue the analogy.
Does pain produce muscle memory? I haven't heard that before.
Maybe a popup would be better, but until we can hack our brains the question remains whether you're better off with shame as a learning mechanism.
Does pain produce learning? Probably yes.
http://www.nytimes.com/1991/08/27/news/pain-and-learning-may-be-close-cousins-in-chain-of-evolution.html?pagewanted=all
I disagree--perverse incentives. If I was ashamed every time I was wrong, I might be more careful. Or, I might stop admitting I was wrong. I make an effort to congratulate myself for admitting I'm wrong, for this reason.
Now that I think about it, it would be even more helpful to find out which of these I would do.
Actually, "No valid inference."
A man who has admitted in the wrong may or may not be thought much of by Abraham Lincoln is the only thing we can infer, and it's too indefinite to count as a useful deductive truth.