Monday, February 1, 2010

The delusion of optimality

In life we usually have to decide. We got to make some assesment of the overall picture and get to the better (optimal) way of doing.

This is how life goes, and I have no argument against it

Yet people tend to get caught in what seems to be better and assume it is an absolute better. Assume the reasons for chooosing were 100% relevant etc.

So, supporters of democracy beleive that democracy is the ultimate way of managing a soceity, it is only good, it is a moral obligation etc.
The reality is that democracy has many failures, its moral value is limited to the situations it is good for soceity, and its betterness depends on context.

Same for experts. Experts have hte advantage of being immersed in a topic and so on. Yet their mind has its limitations + the very disadvantage of being an expert (you program your mind in a certain way and you stop seeing things hte average persons sees).

Same for optimization in general and for science.

Do people want absolutes? do we need something to cling on? Is the value of eleiving worth its drawbacks (in terms of results, not that of process. I do not beleive truth has any inherent value. Its only the value of relating to reality that is sometimes useful)


Anonymous said...

Karl popper said we are better off if we use the 'piece meal engineering' in order to optimize life, but Taleb goes further saying, we can optimize life studying the unseen, what is outside our scope or the ordinary. An example would be in a restaurant bussiness, you can optimize this bussiness studying markets, accounting, recipies,management, statistics, etc, but what is outside this is more consequential like a drop of prices in meat, a hand gun robber, a fire, a suit from a costumer,and earthquake, a flood, an owner hearth attack, etc. And usually these things are outside the entrepeneur and even or worst from the proffesional in bussiness, they might take this things like superstitions thoughts. Greetings from Mexico.

Yechezkel Zilber said...

Piecemeal engineering is another method, which works at times, but it still relates to the idea of op-down management. Taleb does not like top-down too much, but top down has its virtues.

I do not know

Anonymous said...

HI, I don't think these solutions are Top-down, the inner concept would be the 'I don't know' what will happen? It can happen the worst but the solution would be insured or be conscious of the risks.

The new security box or word verification in your blog is an example, in this case you are using the pice meal engineering, but you could use the unseen study or study all the risk can bear in a blogg etc etc.

spldbch said...

I don't like when people believe they are the keepers of the "absolute" truth. Is there really any such thing?

Anonymous said...

I don't believe in relatives or aboslutes truths, saying that somebody has absolute truth is meaningless, because in reality 'there is truth or lie', those whose theorize with out empirical facts are academics Top-down.

It is a joke hear somebody from a University like my sister, saying that all truths are relatives or there is not truth just halfs truths, but they don't realize that those statements are self-defeating.

i.c.hir said...

my first comment apparently got lost because I didn't notice the letters, now I don't remember all of it...

but one question was: what kind of truth is "optimal" or "working", is it in any optimal or good to believe in facts and / or truth?
but probably the better question is in what facts one should believe (as one cannot know all the facts anyway)

another aspect was optimization and facts:
the book "Fundamentals of Piano Practice" ( claims to be the "best" - and I'm convinced to a large extent. There are a lot of cool ideas on optimizing (though sometimes specific to the example of piano). Obviously this is a quite "well-defined" problem (though it gets fuzzy when musicality comes in), thus favoring "optimizing" and "scientific methods"...
So is this valuable?

In the end I don't care about it, I'm just happy about reading some cool ideas (that's also the reason why I follow the blog, by the way), which may sometime, somewhere pop up and help - or never.

Anonymous said...

Trial and error or piece meal engineering to seek truth, you can see lots of whiteswans and you realize that all are white, these are facts, but what you have found is the so called 'problem of induction' How many we have to see in order to stablish it as a truth.(see hume). But one single black bird can make the difference, and this is truth. Empirical facts are necesary but you have to be skeptical, then karl popper came with the epistemological solution or falsacionism tecnique. You don't know beforhand what is truth but you have the choice to find it, lack of curiosity or erudition make people so blind that many are followers believing the shallow part of the structure. How do we know if somebody love us? Facts are necesary but being skeptical all the life can be sad, till you die you may know, this is hard and ironically just falsacionism can help, this is not like 2+2=4, or the sun is the center of the Universe. But remember when Galileo's time people believed thet earth was the center? You see we are shallow by nature and facts are not necesary truths.

Yechezkel Zilber said...


the idea of optimality has uses etc. just a little naive to overstate it. Especially when people gert caught up in the practical general approach and holify it (thinking that democracy is asbolutely better in all circumstances + making it a vaLUE ON ITS OWN + + etc. etc.)

Yechezkel Zilber said...

see also taleb's blog piece 111

"People think that I wrote TBS to communicate my ideas about human errors, epistemic arrogance, complexity, and high-impact uncertainty. The fools. I wrote a book to talk about Yevgenia, Lebanon, Casanova; I wanted to express my love for il Deserto and my outrage for the very existence of frauds like Robert Merton le petit. And I used that Black Swan idea as an excuse. Any other topic would have bored me. Had I written a book about the black swan idea almost nobody would have read it.

Some people think they attend the opera for the story.

It is the same with language. Language is largely made to show-off, gossip, confuse people, delude them, charm them, seduce them, scare them, exploit them, etc. And, as a side effect, convey information. Just a side effect, you fools.

Anonymous said...

I read his blog, and when I read it, I wonder if Taleb knew that The blackswan book was written for both sides. I meant to say to academics and laymans. Academics can learn from Yevgenia, Casanova and Fat tony, and laymans can learn from Tetlock,Kahneman,Popper,Mandelbrot etc. I think that complexity isn't bad at all, we need complexity when we have simplicity. And simplicity when we have complexity, to tell somebody that the milk comes from a cow and is white is not enought, we need some complexity to know the truth. I pick the Blackswan because I'm a currency trader, and I did learn a lot, but I did discover a lot too and I thank Taleb to share his knowledge. Many people got bored with the Blackswan because he hasn't show the perfect formula to succeed but they are blind or shallow, you will even need curiosity and erudition to take the best part of Blackswan Idea, We are not rationals in decision making.

Yechezkel Zilber said...


I agree that one needs a brain to appreciate taleb. I remmeber observing that the reviews in amazon on fooled by randomness were almost binary, those who said "brilliant etc." and those who said "we learned it in school", I gathered that one needs a little bit of a brain to get the ideas.

i.c.hir said...

I surely not the one who will defend "optimization" (as in any sense optimal), I guess I believe more in "improving" if improving comes along easily...

that also brings along that I don't try hard to "improve" on the truths of others (but accept that they are more often than not different, without doubting myself too much)