Saturday, May 9, 2009

The nature of Probability

There is no absolute data. Our eyes can only see so far, our memories can only include our own experiences, our minds can only predict so well. Our instruments are only somewhat superior- Heisenberg's uncertainty principle tears at our best microscopes, and even our best supercolliders can only break down matter so far. Furthermore, we face the threat that not all effects have causes. Probability is the methodology of building upon information. We build up large sets of facts, and we look for congruency. Once we find congruent elements, we turn and look for holes- and start filling them. We've seen many people throw a baseball- if we notice someone starting to throw a baseball, then turn away, we can assume that the baseball will sail off like all the others, even though we aren't looking. However, not every time that a baseball has been thrown, has it sailed off. Sometimes it has gone backwards, or just fallen to the ground. Thus do we create probabilities. We need not declare something so prone to error as "the baseball sailed into the sky." We can simply state "It likely sailed off. Other likely candidates include a foul or a ball." In every situation there is a chance of this scenario- "Something never seen before happened." A universe is but a program, a simple set of rules and data set in motion. As such the rules can be anything, and could always include a line stating something like "if ball is thrown let it soar if it is not the third day of June 2021, else explode sun." our ability to correlate events is limited. The probability that a given event will not, in its final shape, correlate with any other known event, is chaos.

Due to the nature of our universe, and ourselves, we are behooven to act without absolute data. To change the world without knowing what we are changing it from or to. We measure events based on value and act in ways that ascribe high probabilities to good things and low probability to evil things. The chances of a meteor hitting the Earth and killing us all is pretty low, but this event would be very bad, and as such deserves far more of our power than a random terrorist cell that could, if it got very lucky, destroy a singular building. In the same way, some things can be highly unlikely to succeed, but very good if they do. A research project designed to overturn the conventional laws of physics and come up with a perpetual motion machine may well be worth funding. (yes, I know, such a project is probably a fraud, but imagine it was headed by Stephen Hawking or some such trustworthy individual. On an unrelated note, the act of giving geniuses money to do stuff that doesn't make sense is often a good idea. Even if you yourself are a genius, another genius may well have thought of something you haven't and come up with an idea that, ingenious as it is, doesn't make sense to anyone but themselves.) Since we can never truly be sure of what we are doing, it is a good idea to allow for backup plans, and alternate roads to success. IE, to allow for different worldviews to exist. Anyone who is actively hurting our plans is in our way and must be destroyed, but people who are even slightly profitable to our overall schemes are valuable fail safes. As such those who argue against us are not in our way (unless they are calling for the destruction of ourselves or our ideals.) since others can either ignore them (and thus support us.) or agree with them (and thus turn into fail safes.) Good ideas, once implemented have geometric value. If an idea is good, then evolution will join us, and spread that idea's champions far and wide. If an idea, once implemented, fails to succeed, one should not assume that its failure was due to outsiders. In particular, a inferior group will never defeat a superior group without the support of the superior group. It is when a superior group degrades itself to the level of its inferiors, that the barbarians are able to storm the gates, and provide the final rites for their once-great enemies.

America does not have to tolerate corruption. I see countless arguments flying back and forth over whether 9-11 was an inside job, or whether Obama is legally allowed to be president. That's looking at it wrong. There are millions of people who are just as qualified to be high up politicians as the ones we have today. If a politician could be corrupt, if we can come up with the slightest evidence that a politician has something wrong with him, he should be replaced. it costs us nothing to find someone who is just like him in every way, except that the new person has a pristine track record. We act like removing a politician from his office is equivalent to executing a pedestrian. It's not at all the same. When a politician is removed from office, his wealth does not vanish, nor does the food and comfort that wealth provided disappear. Furthermore, far from his time dispersing, he is freed up to partake in other valuable activities, such as running businesses or making movies. As for finding his replacement- we have, today, created a society that knows a ton about a multitude of its people. If you've created a successful business, we know. If you've committed a crime, we know. From there we could easily generate a list of people who pass corruption and effectiveness muster, and are approved by their constituent parties, then hold a quick vote (which should be moved purely online btw.) and put in a new person. We could have actual turnover in politics, instead of the modern system where those in power stay there, free to become steadily more corrupt indefinitely.

Remember, our goal is to act in such way as is likely to make the universe better. No one else has any hold on us, there is no obligation, there is no responsibility. There is only good. We are not obligated to buy our medical services from those who have gone through med school or to work for someone who has a large pool of experience. We see only the past, live only in the present, and can only affect the future. As such, we must judge based on the past, act within the present, and place our goals in the future. Thus the present supplies the eternal pivot point, the location from which we decide what to do. Regardless of the past, it is used only as a well of information, it does not control the present.

Track records provide a good correlation between a person and how well an investment in the person will pan out- but it is not the only information that provides such a correlation. In the past we relied on genetic descent to judge such problems. This worked because genetics has a high correlation with intelligence. Today, we have a more accurate method that pinpoints the particulars of the problem- IQ. With our technology we can find out exactly who is best suited to a given task, and act accordingly. The largest barrier to this approach is experience, but the answer to that is specialization. If one person handles all the appendicitis then they will be handled extremely well. If it is necessary that a new person starts doing that job, he should be introduced as an apprentice for a number of years. Brute knowledge, a issue that was of high importance in the past, is no longer important. This is due to computers, databases and the Internet. A smart doctor who has been working for 2 years will be able to diagnose someone just as well in front of a good database+a search engine, as a equivalent doctor who has memorized every sign and symptom. Once a person learns how to find information, it is as good as his. The beauty of relying on tests and other scientific measurements to find out a person's capacity isn't just that we can skip educating people or that we can scientifically search out the best people for the job. It's that we can put the right people into positions that are hard to judge. If a person breaks his machines on an assembly line, or writes faulty code, it's pretty obvious that he is doing a bad job. But what of a mathematician who goes to work every day, tinkers with Fermat's last Theorem, achieves nothing, then goes home? What of the politician who goes up, passes random bills that may hurt or help the nation, then leaves? What of the engineer who stands vigilant over a nuclear reactor that hasn't ever actually broken? Abstract testing can discover in ways that credentials and on the job assessments can't, how well a person is actually doing a job.

No comments:

Post a Comment