Slashdot Mirror
Why Being Wrong Makes Humans So Smart
Hugh Pickens sends in an excerpt in last week's Boston Globe from Kathryn Schulz's book Being Wrong: Adventures in the Margin of Error. "The more scientists understand about cognitive functioning, the more it becomes clear that our capacity to make mistakes is utterly inextricable from what makes the human brain so swift, adaptable, and intelligent. Rather than treating errors like the bedbugs of the intellect — an appalling and embarrassing nuisance we try to pretend out of existence — we need to recognize that human fallibility is part and parcel of human brilliance. Neuroscientists increasingly think that inductive reasoning undergirds virtually all of human cognition. Humans use inductive reasoning to learn language, organize the world into meaningful categories, and grasp the relationship between cause and effect. Thanks to inductive reasoning, we are able to form nearly instantaneous beliefs and take action accordingly. However, Schulz writes, 'The distinctive thing about inductive reasoning is that it generates conclusions that aren't necessarily true. They are, instead, probabilistically true — which means they are possibly false.' Schulz recommends that we respond to the mistakes (or putative mistakes) of those around us with empathy and generosity and demand that our business and political leaders acknowledge and redress their errors rather than ignoring or denying them. 'Once we recognize that we do not err out of laziness, stupidity, or evil intent, we can liberate ourselves from the impossible burden of trying to be permanently right. We can take seriously the proposition that we could be in error, without deeming ourselves idiotic or unworthy.'"
Interesting way of looking at our failures. So... let's see if BP uses this to prove their genius.
Once we recognize that we do not err out of laziness, stupidity, or evil intent, we can liberate ourselves from the impossible burden of trying to be permanently right.
Sometimes people do "err" out of laziness, stupidity of evil intent!
We can take seriously the proposition that we could be in error, without deeming ourselves idiotic or unworthy
Any suitably intelligent person already knows that failures are as much a part of learning as always being "right". And sometimes we do make really silly mistakes by overlooking things that should have been obvious. I know I do. Then again, often what is obvious to me, isn't to others..
which is totally what she said
I'm never wrong.
I thought I was once, but it turns out I wasn't.
I'm sure we've all noticed that the people who make the biggest mistakes get promoted the fastest.
If you were blocking sigs, you wouldn't have to read this.
I think people focus their criticism more on those that make errors that seem glaringly obvious to everyone else. We tend to call those "stupid" errors. It's true however people tend to become far too critical of others who seem to be unable to reach the same conclusions at a high speed that we have already come to.
On the other hand, there are obvious mistakes that should not be conflated with probabilistic errors due to inductive reasoning. When the heads of BP cut corners that result in a giant explosion, a several month long oil leak, and billions of dollars in damage to the environment and people's lives, we can attribute that to gross negligence.
When a politician decides to engage in 2 costly wars while lowering taxes for the rich, or when a majority of society elects politicians who repeatedly punish the poor and middle class while rewarding the rich, and then complain about not having enough money to support their expensive lifestyles, you can attribute that to stupidity.
The funny thing about that is your post wasn't all that funny. So you're even more wrong than you think.
I have known this for most of my life. The name reflects the idea. I'm not afraid of being wrong... at least not as much as others seem to be.
The depth of the value of errors goes much further than the topic describes. The animal brain itself is a noisy collection of errors. The reason correct processing happens at all is because nearly all possibilities are explored in neural pathways to get to the correct responses. Once correct responses are identified, neural pathways to the correct response are established. This is what we call learning in the lowest level sense of the word.
I have always found it amusing and interesting that computers work the way they do. They work in ways that are the complete opposite of the animal neuromechanism. Computers, originally derived from numerical processing devices, rely on accuracy and seek to prevent errors in every way possible. Memory is storage rather than a path. In a way, computers are our biggest hangups about being wrong put into mechanical practice.
I find it to be far from ironic that we are now trying to get computers to "learn" under these conditions. The fact that it doesn't work particularly well. When every measure is taken to always be right, how can a machine learn? It is also far from surprising to me to see that people who are so afraid of being wrong are also the least capable of learning anything new or useful or being able to adapt to new circumstances. It all fits neatly within my own observations about mistakes and learning.
Mistakes can cost us time and money, expose us to danger or inflict harm on others, and erode the trust extended to us by our community.
Or being ridiculed and humiliated by assholes who gain a false sense of superiority by belittling people over mistakes - many times trivial ones. Which then leads the other person to dig their heals in, argue pedantic points to stay "right" which then leads to counter pedantic arguments from the other, and round and round we go!
But hey! That's what you get when you post on Slashdot or work in IT.
RIP America
July 4, 1776 - September 11, 2001
While it might be true that "we do not err out of laziness, stupidity, or evil intent", it is definitely also the case that laziness can and does lead to ignoring procedural correctness that would have caught error, stupidity can and does delay the recognition of error until it has had time to balloon into something more serious, and evil intent can cause the willfull application of anything that laziness or stupidity would lead to; but carried on much more intelligently(and thus dangerously). Not to mention, of course, that little class of statements we know as "lies", which are essentially calculated to cause errors in those receiving them.
Obviously, in a trivial sense, nobody wakes up in the morning and says "Gosh, I sure do feel like really fucking up today!"; but some people take measures that reduce the probability of error(and, where possible, measure it) and others do not. Just because virtually all human reasoning, outside of (some) math and syllogisms, is inductive does not imply that all human reasoning is on equally firm ground. In fact, given that deductive logic is useful pretty much only in certain types of math and in carefully controlled toy situations, the ability to distinguish various statements of inductive logic by quality or probability is probably the most vital aspect of epistemology as an applied science...
To #ERR is human, to forgive divine.
http://alternatives.rzero.com/
that we could be in error, without deeming ourselves idiotic or unworthy."
i guess Schulz has never read a comment board
intellectual property law is philosophically incoherent. it is your moral duty to ignore it or sabotage it
If loving you is wrong, I don't want to be right!
"The average reporter we talk to is 27 years old......They literally know nothing." - Ben Rhodes
So that's why I feel smarter after staying at a Holiday Inn.
None of these conclusions make sense in an Eastern shame culture/honor culture. These conclusions, do, however, dovetail nicely with Western guilt culture. Correctly pointing out the mistakes of others can result in massive loss of face for the correctee. This will have real consequences for the finger-pointer. Publically admitting that you were wrong and redressing your errors is career suicide in many places throughout the world. I see it all the time, Westerners are shocked that their culture of "it's OK to make mistakes and it's a positive thing to admit when you are wrong" doesn't apply everywhere.
Shutting down free speech with violence isn't fighting fascism. It IS fascism!
Look at what happens in Japan when a major mistake is make and in the west. Has anyone from BP taken accountability? Has anyone from Boeing ever laid down their jobs because they killed a couple of hundred people with their bad decision? Has any airline director every left? No.
But in Japan the higher ups DO feel that they are at fault for mistakes.
Your explenation of western attitude often becomes: A fault is nobodies fault.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
http://en.wikipedia.org/wiki/Bloom_filter
...for Lucy is never wrong. (There is some kind of circular logic there...pumpkin-shaped, possibly.)
Orwell: "In a Time of Universal Deceit, telling the Truth is a Revolutionary Act"
"There is no such thing as an absolute truth."
Is the above statement absolutely true?
and they were right, just look at how many times he was wrong!
Monstar L
Or deliberately ignoring your own engineers saying, "This is a bad idea. The wellhead will blow out." Then try to act all surprised to discover the engineers knew what they were talking about, and blame the engineers instead of your own stupidity Mr. BP Manager.
"I disapprove of what you say, but I will defend to the death your right to say it." - historian Evelyn Beatrice Hall
Often, the only way to get answers to your questions on the internet is to claim things about the subject you know are wrong. Then heaps of people will jump on you to tell you what is correct.
My other account has a 3-digit UID.
Shulz is precise, just not quite accurate in her descriptions, assertions and conclusions.
It's not (just) inductive reasoning that produces the humans' results, it's heuristics. We create the fastest good enough result rather than the best possible result more slowly. The former proved conclusions that are correct enough but very fast, which evolution favors over slower but more accurate decision making. You can be right as god, but if you get ate you're just very right poop.
Heuristics works in all directions, top-down, bottom-up and side-to-side. Inductive, deductive and all the rest is labels we developed much later to try to describe what we could figure out about what's really going on in our heads. We can do those things because they're all part of how we work, but on the fly we never work in only one direction. Heuristics develops chains of thought according to associations, and so can fill in the chain (more often, the tree)
There are some things that defy logical reasoning, such as language. We can use reasoning to figure out how to talk about the arrangement in memory of the items we can recall and so talk about, but learning to communicate happens far faster than learning can account for. Hence "generative grammar" and the utterly arbitrary nature of language production. Such things are predetermined in the way of species specific behaviors. We are genetically predisposed for these, and no logic could possibly keep up. This could be hardwired heuristics, though nobody can prove that as yet, but it certainly acts like it.
So, heuristics, not induction, plus hardwired exceptions. Thus, we're never right, but we're right enough (to varying degrees) fast enough to survive.
Top Shulz's cake with that frosting, and her precision becomes accurate also when it comes to our (neuroscientists) present best picture of how we think.
It's not in the article above, but thinking that's always completely right has the major failing of being unable to produce novel responses. Heuristics allow the adaptability which novel situations require (another ability favored by evolution as well as Dr. Chandra), and which allows for creativity.
Sounds like a very good book. Adequately correct too. Must have been written heuristically.
"I may be synthetic, but I'm not stupid." -- Bishop 341-B
David Hume pointed all of this out hundreds of years ago. And he backed up all his claims with plenty of evidence that was readily available at the time.
I wonder if Kathryn Schulz's is aware of this?
Why is the "BP manager" currently out on a yacht at some annual event instead of sat in court, desperately defending himself from a public prosecutor with a battery of lawyers funded by the US government, WWF, and GreenPeace? No doubt they all want a piece of his personal fortune... Especially the lawyers.
He shouldn't be out sailing, he should be taking a plea bargain involving a few hundred million dollars in personal fines and 15 years in a federal prison.
Finally had enough. Come see us over at https://soylentnews.org/
we need to recognize that human fallibility is part and parcel of human brilliance
Yet another reason why the idiots pushing for instant replays in baseball should STFU.
I think that anyone who has dabbled in machine learning would not be too shocked (weather by Hume's version or this post). It's the error term in machine learning, adaptive filtering, etc. that really drives the learning. As a stupid but simple example: Least Mean Squares in adaptive filtering (essentially gradient descent over the error surface).
@humanity: *facepalm*
Isn't that what management does best? Takes credit for success, and passes blame for failure? It's the only way to get into the Fortune 100 C*O offices that I'm aware of.
My blog. Good stuff (when I remember to update it). Read it.
DEDUCTION: Rule + Case -> Conclusion
Induction and Abduction use the elements in a different way:
INDUCTION: Case + Conclusion -> Rule
ABDUCTION: Conclusion + Rule -> Case
Only deduction provides a valid inference. But humans default to using abduction and learn induction and deduction only slowly through formal training.
Due process, or some other such legal technicality.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
I'm from Scandinavia, one of the biggest cultural differences that I notice in my day-to-day work with people from other cultures is that people from other cultures tend to have a lot harder time admitting that they were wrong. They see this as some kind of defeat, and the really stupid ones keep pushing their idea even though they know it's bad. Here, people won't look down on people admitting that they were wrong, instead you will get "respect" for acknowledging that you were wrong and taking corrective action. This kind of open attitude is a necessity if you want to see innovation instead of fear...
Why is the "BP manager" currently out on a yacht at some annual event instead of sat in court, desperately defending himself
Same reason the Union Carbide guy, who killed tens of thousands of people in Bohpal, is living off his life in luxury in the U.S: The system is made by the rich. for the rich.
You can't take the sky from me...
Kathryn Schulz's book makes a great case for understanding why being wrong is so intrinsic to being human...unfortunately, and ironically, she's got it 180-degrees-wrong.
Where she fails is her conclusion: it's not that BEING WRONG is what makes us so successful, adaptive, and smart. It's the 'trying again to be right' bit.
Being wrong is easy. Being right is much, much harder, and probably requires trial and error. But if you're satisfied with being wrong, you don't keep trying. While the idea that 'being wrong is human' is all nice and friendly, ACCEPTING being wrong without any sense of negative consequence is staggeringly, blindingly stupid. Without gradations of consequence (ie more and more serious consequences for more and more serious failures), life doesn't even make sense.
"Schulz recommends that we respond to the mistakes (or putative mistakes) of those around us with empathy and generosity and demand that our business and political leaders acknowledge and redress their errors rather than ignoring or denying them. "
Sorry, but that's just stupid. This is the same sort of touchy-feely crap that's infected modern American public schools. "It's ok, little Timmy, you just keep trying to figure out what 2+2 is. You're still a valuable and precious little snowflake."
Why should Timmy ever bother to figure out 2+2 if he never NEEDS to get it right? Whether it's reward-based or something more simple like shame, there MUST be a disincentive to be wrong. Anything else is simply asinine.
So you send your husband out to get dinner; instead of buying food for your children, he spends the money on porn and beer. Ah well, you should respond with generosity and empathy, right?
Can you imagine if her methodology was followed? "It's ok BP, we all know that drilling for oil is hard work, and can "It's ok, Mr President. You just spent well over a $trillion on an ostensible economic rescue plan, but aside from simply not working, it pretty much all ended up in your friends' and political allies' pockets. We won't be angry, we won't even be annoyed. We'll respond with generosity and empathy. Perhaps you could take another $trillion from our kids' and grandkids' future and try again? Maybe this time you'll succeed?"
-Styopa
Or deliberately ignoring your own engineers saying, "This is a bad idea. The wellhead will blow out."
Nobody said that, and if they thought it they had the authority and the duty to stop the operation. I work as a contractor for BP, and they pound it into your head over and over that everyone has the authority and obligation to stop a job if they think it is unsafe. It is one of BP's eight "Golden Rules" of safety. Everyone on-site - BP employee or not - has this authority and duty, it is a condition of employment for BP and all its contractors. If there were engineers who believed the wellhead would blow out because of the course they were taking, they should be held liable for the deaths of their coworkers, because it was their job to stop it, especially if management thought the job was safe.
To be clear, blow outs happen. They are a fact of life in the oil industry, and to think you will be able to prevent them 100% of the time is idiotic. From what I've heard so far, most everything that happened on that rig was within industry standards, and while hindsight makes it clear there were some serious mistakes there, those mistakes were not at all obvious at the time.
That's why they invented Blow-Out Preventors, they are specifically designed to prevent exactly this type of catastrophe, and they are installed on every single well in the gulf (and any other off-shore rig). This is where the real problem happened. It seems that the combination of pipe and BOP were not conducive to actually sealing the leak, and this is a serious error. The cost issue is somewhat of a red herring. The three easiest ways to get funding at BP for a project are safety and compliance issues, environmental issues, and production issues, in that order. Among these, BP will try to get the most "bang for their buck" on any given project. This usually means completing the task at the lowest possible cost. That's where it bit them this time - the low cost option is normally fine, but obviously under 5,000 feet it is not acceptable. That was not known before hand (though most companies do go with the more expensive option in this case, just to be safe), and in fact the US Government signed off on everything BP, Anadarko, Haliburton, and TransOcean did every step of the way. BP did nothing without approval from regulators, which is how all oil fields operate. Everything must be in compliance, and everything at the DeepWater rig was (at least according to MMS at the time).
As always happens after a catastrophe, industry standards will be changed, and the initial blow out will be less likely. This will always be their though, and the BOP's are designed to stop that.
Security is mostly a superstition... Avoiding danger is no safer in the long run than outright exposure. - Helen Keller
Well, in theory that is the system that is in place in most Western countries. I'd venture a guess that in most cases, it's not in the interests of the people in charge to have justice run its course properly.
The other thing that one would need to consider is, despite what everyone said, who is at fault? Was it an accident? Or was it on account of negligence or evil intent? (Or stupidity as the article says...)
Not just referring to the BP case specifically there, but in general. Things like that are IMO difficult to determine conclusively.
One thing I know, and that is that I am ignorant...
Given everything I'm hearing about BP's higher than average safety violation problems, I know they may talk the talk, but apparently they do not walk the walk.
Companies SAY things all the time that they really do not mean.
For example at my company we have three status rankings for projects.
GREEN
YELLOW
RED
In three years, for projects which were cancelled, which were late, which were horrible failures, guess which status rankings were NEVER USED.
I foolishly used yellow once and the reaction was strong. I said it was yellow because it was going to miss its date and needed more resources.
Not a good idea.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
As an engineer, I would honestly rather upper management to be as far away as possible, that's how the real work gets done. Show me an engineer that wants a CEO breathing down their neck and I'll show you an average engineer that wants to brown nose with management. Also what is a CEO supposed to do? What in his background would leave you to believe that other than signing 20 BILLION dollars into escrow for repairs/claims that he would be more effective at the scene? I'm not a fan of big business, but people are just looking for a reason to crucify him. I don't go to BP, that's what I do to show my disapproval.
You told me, but you didn't convince me.
Actually, had they told anybody, the job would stop. Every employee has the authority to stop a job - any job. There aren't some jobs that some people can stop and some jobs that other people can stop, anybody can stop a job for safety on a BP rig (or any BP facility). That gets pounded into your head day from day 1 - if you see something that you think is unsafe, you stop it, and everybody gets together and double-checks the plan and makes sure they haven't missed anything that would make it unsafe.
There are practical limits, of course. For example, if I'm not involved in a job and I have no idea if it's safe or not because I'm not qualified enough to know the difference, then I have no business stopping a job. I still have the authority to stop it, but I won't stop a job because I have no idea what's involved. However if I'm involved in a job and I feel unsafe, I will absolutely stop the job.
By the same token, management may be pushing to get a job done a certain way (they always want to use the low cost option), but if they aren't qualified to know what is safe and what isn't they obviously aren't going to stop the job for safety. However, if you are qualified to know if it's safe, and you think it is not safe, you MUST stop the job. If you're working on a job and you feel unsafe, you MUST stop the job.
All it took was for one person to say "This doesn't seem safe, we need to stop the job" and the job would have stopped right then and there. The fact that it didn't means either nobody said to stop the job, or there was a serious breach of BP policy.
In other words, all of this "If they had just listened to the engineers" stuff is either complete bullshit (as in, never happened), or criminal mis-management at the rig level. This is not the kind of decision that happens further up the chain. There is a very real possibility that there was a local culture to ignore safety concerns in spite of BP policy, in which case the ones responsible actually are the people on the rig. Not Tony Hayward, not the President of BP Americas, but the rig management and possibly one level above them (if only for putting such people in a position of authority).
I do think there is a real problem with BP's management culture which makes accidents more likely. They have a tendancy to move managers around from position to position, and they tend to stay at one place for no more than two years. The idea is to get a "broad understanding" of oil field operations as well as the corporate side. This means if they are ever going to get a top-level manager, they can't keep them in one place for very long. This leads to serious inconsistencies in management of a particular facility/rig. They also tie bonuses directly to how much of your budget was left over each year. This creates a perfect storm for accidents due to poor maintenance, as the easiest place to cut is the maintenance budget (safety & compliance and production always gets funded). I believe this is why BP has the worst record for environmental accidents in the industry by a huge margin. How that directly relates to this spill is going to be subtle, though. I would definitely name it as a contributing factor.
Security is mostly a superstition... Avoiding danger is no safer in the long run than outright exposure. - Helen Keller
Or deliberately ignoring your own engineers saying, "This is a bad idea. The wellhead will blow out."
If there were engineers who believed the wellhead would blow out because of the course they were taking, they should be held liable for the deaths of their coworkers, because it was their job to stop it, especially if management thought the job was safe.
Hold it. It was management who was pushing pushing pushing to get that well pumping ASAP, and management who told operators that 2 instead of 3 concrete plugs would be sufficient. It as also management who did not ensure both batteries in the BOP were functional/charged. For you to throw this all on engineers when there are numerous reports of management forcing an unsafely accelerated schedule is ludicrous and shows that you are less than impartial on the topic.
To be clear, blow outs happen.
To be clear: blow outs can be prevented if standard safety procedures are not bypassed.
That is where I take issue with the claims in the parent article. It assumes all humans are interested in being intelligent and learning from mistakes. That is far too optimistic a view. The article actually says 'Once we recognize that we do not err out of laziness, stupidity, or evil intent...' But people DO err out of those reasons (I equate greed with 'evil intent' when the person knows their actions has a significant likelihood to harm/kill others, which is exactly what happened in BP's case.) It would be a major mistake to assume nobody in the future will put greed ahead of safety and make a mistake via that incorrect choice. This repeating pattern is not a sign of intelligence.
"We can categorically state we have not released man-eating badgers into the area." - UK military spokesman, July 2007
The disconnect between this concept and most people's thinking explains why scientists and engineers rarely advance well into management and politics.
- As a scientist or engineer, it is acceptable - even required! - to incorporate new data and adapt your thinking, even reach different conclusions.
- As a manager or politician, such behavior reflects weakness or lack of principle, sometimes called "flip-flopping".
In my experience the latter approach seems to be the *typical* perspective of normal people (non-engineers), who would rather "stay the course" and "finish what they started" even when they openly admit that they would have chosen differently now. The contrapositive concept of "what did he know and when did he know it", with the understanding that someone who chose badly may have made a reasonable decision based on the information available AT THE TIME, is often displayed pro forma and then trampled upon.
Actually, had they told anybody, the job would stop. Every employee has the authority to stop a job - any job. There aren't some jobs that some people can stop and some jobs that other people can stop, anybody can stop a job for safety on a BP rig (or any BP facility). That gets pounded into your head day from day 1 - if you see something that you think is unsafe, you stop it, and everybody gets together and double-checks the plan and makes sure they haven't missed anything that would make it unsafe.
Heh; yeah; that's the official policy in lots of companies. But I've worked a number of places where, when I asked around to find the people who had done that, I quickly learned that those people no longer worked there. It doesn't take a genius to make the right inference from this.
It also doesn't take a genius to understand that if something does go wrong, you were present, you'll be one of the people taking the blame for the problem.
The old-timers just grin and say something like "So you've finally figured out how it all works around here."
Those who do study history are doomed to stand helplessly by while everyone else repeats it.