Monday, June 25, 2007

Fallibility

Fallibility
As a thinking machine, the human brain leaves a lot to be desired. Living is about making decisions, most of them binary choices. Do I go left or right? Do I eat this plant or not? Do I release my arrow now or do I wait for a better shot? We seem to have evolved special mechanisms for making this type of survival related decision. Those mechanisms often get in the way of complex thinking. In fact, I'll propose to you that what we call thinking is only an artifact of our primitive decision making skills, creating the illusion that we can reason. Our ability is so faulty that large fractions of the population, given the same facts about a situation will still come to quite different conclusions as to what to decide. That's why on almost every public policy issue, the population will split very roughly in half, and each half will think the other is mad. With Respect -Joel

9 comments:

Ira said...

Thanks for posting another new main topic.

A dichotomy is a two-way choice. As you point out, many time-critical human choices are binary. Our brains are adapted to making these kinds of choices with a minimum of "thinking." Animal survival depends upon making (mostly) correct binary choices fast.

One of my favorite grad school professors, Walt Lowen, wrote a book, "Dichotomies of the Mind." He claims (with indisputable logic and justification) there are only *two* kinds of people: 1) those who believe there are only two kinds of people, and 2) those who do not!

Our brains consist of billions of specialized cells called neurons. While each neuron receives and weighs electrical signal inputs from thousands of others it can only do only one two things at any instant: 1) fire and send out a pulse of electricity or 2) not fire. Our "thinking" consists of the net result of billions of neuron firings over periods of time.

Similarly, in a digital computer, each memory cell can hold only a "0" or a "1", another dichotomy. Out of all these "0" and "1" processes, the computer does all of its work.

On the other hand, one of the things that most distinguishes primates from the "lower" animals is their ability to do long-term planning (in addition to the short-term quick binary decisions).

Humans are the primates with the best-developed abilities in that category, which makes us the "thinking animal".

I would dispute the conclusion of your posting to some extent. It is true that the public policy *issues* are those on which we are nearly evenly divided. However, that is *why* they are issues - *because* of a nearly equal divide.

The majority of public decisions are less controversial because they divide very unevenly, with 60 or 70 or 80 or a higher percentage of people agreeing. The small minority that disagrees usually goes along and accommodates the majority, even as they maintain their disagreement.

An exception is the extremists, such as the Islamic terrorists. They are so dedicated to their minority opinions they are willing to sacrifice their lives killing innocents in an attempt to cow the majority into fearful inaction or surrender.

joel said...

I'm not so sure that the majority of decisions are less controversial. One of the things that divide C-Minds and L-Minds is social versus individualist thinking. Thus things as simple as whether or not to add a traffic light at an intersection where there have been traffic deaths can be at issue. Traffic flow will be impeded. Lives will be saved. How are they to be balanced? When there is an overwhelming majority in favor of a proposal, I think it's because of a lack of presentation of all of the arguments. with respect -Joel

Ira said...

Joel: I'm trying to make sence of the Conservative-Mind vs Liberal-mind and Social vs Individualist thinking dichotomies as applied to the traffic light example you brought up.

Here are two arguments for or against:

1) A traffic light at intersection "A" will cost "X" dollars for installation and maintenance and delay traffic on the main road at an estimated cost of "Y" dollars to the individuals delayed, over a ten year period. A traffic study indicates it may save "Z" lives over that ten year period. Valuing a life according to insurance law at "L" dollars each, just run the numbers and make the decision based cost-benefit, rather than emotion.

2) My neighbor's child lost her life at that intersection. How can you refuse to put a trafic light there based on heartless cost-benefit analysis?

I favor (1). Is that C-mind or L-mind? Is it Social or Individualistic?

Ira Glickstein

joel said...

Ira said:
I favor (1). Is that C-mind or L-mind? Is it Social or Individualistic?
Joel responds:
Yes, Ira. That's an example of the divide between L-minds and C-minds. In terms of individualst versus socialist (not the economic system), the individualist view demands that one examine the loss of freedom and money to the individual when considering the traffic light. The socialist point of view considers what kind of society would allow mere money to come befor the welfare of children. With respect-Joel

Ira said...

Thanks Joel!

So, I score as a C-mind (vs L-mind)and "individualistic" (vs "social") on the traffic light issue. What a relief!

I was concerned that my high consideration for the inconvenience of delays of lots of people would be considered "social" thinking and that the neighbor's high consideration for the life of her neighbor's daughter and the lives of others who might be killed at that intersection would be considered "individualist" thinking.

I would enjoy it if some L-mind "social" thinker reading this thread would give his or her opinion.

QUESTION: Should we put up a trafic light at *any* intersection where we could *potentially* save a life over the next ten years despite the inconvenience and lost time and money for the thousands of drivers who would be delayed by the light?

What is wrong with running the numbers with a cost-benefit study and using that to make the decision? I am in favor of trafic lights where they are justified by cost-benefit analysis.

Ira

Stu Denenberg said...

I am reminded of the recent study of not fully developed teen age brains which explains why they do such " silly teen age things". Surprisingly it's not because they can't reason --- they can do that just fine. The problem seems to be that the mechanisms for making critical decisions without thinking have not yet fully developed. So, for example, in a critical driving situation, they tend to try to think it through when they really need to use another part of their nervous system --- what we tend to call a reflex I think. All this is very counterintuitive...

Regarding the L/C evaluation of the traffic light scenario, I believe there is an old Jewish saying to the effect that saving one life is the same as saving one million and conversely I suppose killing one person is the same as killing one million. This only makes sense if the worth of a single human life is infinite which to most religions it is. I would say that if you could demonstrate that there exists a high probability of another death taking place if we don't build a traffic light (or some equivalently effective solution) at this dangerous intersection then we must do it regardless of the results of any cost-benefit analysis. Does that make me an L-brain?

Stu


Stu

Ira Glickstein said...

Stu:

Thanks for trying to represent the L-mind in the traffic light scenario. I appreciate your effort.

Yes, saving one life is (at least in religious theory) the same as saving millions.

If one followed that literally and completely, there would be a traffic light at every intersection and we would all have armored tanks as cars. There would be a bridge or tunnel for people to cross every street. That would eliminate nearly all traffic accidents!

Continuing that thought process, any product that could conceivably cost a life would be banned. No swimming pools, stoves, electric lights, bicycles, and on and on. We'd have to eat cold, raw food in the dark.

Dangerous jobs would be eliminated. There would be no police (they might lose their life in a conflict with a criminal). No jobs where knives or saws or any sharp tools had to be used (someone might accidently get killed). We'd all live in mud huts and the criminals would rule!

You get the idea!

Such excessive caution would put a drain on our economy. We could not afford any medical care at all and HOW MANY LIVES WOULD THAT COST?

In addition to saving lives, we need to make life worth living. I doubt life in a total "rubber room" world would be worth living.

So (my C-mind insists) it still comes down to a cost-benefit analysis. How many lives must we sacrifice to accidents to save lives using potentially dangerous tools and technology? And, what level of life is not worth living?

In zoos we protect the animals from predators and dangers and feed and house them. Even if it was possible, would we humans want to live in a zoo? If that what L-minds long for, a zoo-like "ideal" existence? If so, who will be the zookeepers?

Ira

joel said...

Ira, it seems to me that you are saying that there is a correct method of making the decision, and that although L-Minds will not accept this method, C-Minds will recognize its correctness. I don't believe this is right, even in a strictly engineering sense.

You have decided to seek the minimal cost solution, by taking into account the legal value of a human life and the expense of the traffic light. I could quibble about the factors to be included. Are you going to consider secondary factors such as the value of the extra time that drivers spend on the road? How about the probability that a Michael Douglas type will get frantic in his car and go out on a vigilante killing binge (what was that movie?) There is no standard for the value of a human life, hence the need for a jury to determine damages after the fact. In courts, we also allow punitive damages and damages for loss of companionship. I could go on this way concerning the practicality of the computation, but there is something much more fundamental than quibbling about the above.

An optimization is not automatically about minimal cost or maximum profit. That's a decision based upon "corporate ethos and the business plan." There is no reason why the function to be optimized cannot be something other than profit or some combination of terms each weighted with a "value factor." One might choose to maximize esthetics or minimize environmental damage or minimize fuel consumption (including funeral processions) or minimize wear and tear on the road surface (including cleaning up splattered pedestrians). My point is that your decision to make monetary expense the function to be optimized is totally arbitrary.

It isn't the solution procedure that separates L-Minds and C-Minds in this case. It is the entirely valid argument over what should be the objective function to be optimized.

Ira Glickstein said...

Joel:

I do not claim C-minds are any better than L-minds in finding the "correct" solution, or event that there is an absolutely correct solution to find.

If we could see 10 years into the future, we would put traffic lights at the exact intersections where people would die absent those lights! We would not put lights at intersections where people would not die. That would be an absolutely correct solution.

The problem is, we can't see 10 years into the future!

Faced with that uncertainty, an "unreasonable" C-mind might do away with traffic lights all together and let the fittest pedestrians survive.

Given the same uncertainty, an "unreasonable" L-mind might put traffic lights (or pedestrian bridges) at all intersections.

If we only include "reasonable" folks, whether they are C-mind or L-mind, we have to do some sort of a value-benefit study.

Yes, they will quibble! An L-mind might put a higher value on human life and aesthetics and a lower value on traffic delay than a C-mind.

So, we will quibble about the details of the value-benefit study and bring up factors like the variability of legal judgements of the value of a life or wasted gas while idling at a light and so on and on.

When the C-mind suggests using a higher life value in prosperous neighborhoods and a lower value in poor neighborhoods (according to the legal judgements in those areas) the L-mind might react emotionally at "unequal treatment". When the L-mind argues environmental impacts and aesthetics the C-mind might get furious.

Yes, the "objective function" as you call it will be a point of contention and I agree monitary cost is not the only factor to be considered. (But money is way ahead of whatever is second :^)

At the end of the day, however, (almost) everyone will agree we need traffic lights at intersections "A", "B", and "C" and we don't need them at "X", "Y", and "Z". Then, we will argue over "L", "M", and "N".

Ira Glickstein