Wednesday, August 18, 2010

Artificial Intelligence

Artificial intelligence has many definitions, some define as "intelligent machines" such as a robot, others define as "the study and design of intelligent agents" where an intelligent agent is a system that perceives its environment and takes actions that maximize its chances of success. No matter what definition you agree with, AI is a reality. Though it is important to make a distinction between computing power and the ability of computers to think, it all comes down to how fast a machine can perform an operation. Everything a computer does breaks down into math. Your computer's processor interprets any command you execute as a series of math problems. Faster processors can handle more calculations per second than slower ones, and they're also better at handling really tough calculations. Computers cannot think (as of now), they process information based on a given set of instructions. So, it is the computing power that has allowed AI to become what it is today.


Analytical Software, Expert Systems, Artificial Neural Networks, these are just a few names that were nonexistent 20 years ago. Today, Artificial Intelligence is a multibillion dollar industry. The uses and applications are widely varied.

Companies are taking advantage of AI and changing the way traditional tasks/methods in many industries are done. Many of us use the internet without realizing the behind the scenes that goes on every time you visit a website. Advertising is the biggest one, pay attention to those ads that are customized for you, many of them even contain your name, so how do they do it? AI and software. Here are a few industries and application that will be implemented in the near future.

  • Sales forecasting

Many companies dedicated to retailing depend on sales forecasts for their success as a company. For them producing accurate forecasts is a critical activity, typical forecasting methods can introduce errors that can result in inefficiently allocating resources or planning, for instance, preventing stock outs is very critical for retailers. With the introduction of Artificial Neural Networks in forecasting the accuracy of its results is significant. Part of this is because predictors use advanced neural network technology which does not suffer from the limitations of traditional methods. In the future companies will not have to worry about past data because with this new technology forecasting will be made in real time that is, these systems will be able to see any changes in the patterns as the sales occur. This will help reduce inventory costs, improve planning and therefore save time and money.

  • Industrial process control

Expert Systems are already in place in certain industries. Expert Systems are widely applied, from sensor validation through production change control, plant-wide optimization, and dynamic scheduling. Combining an ES and ANN technology can be very beneficial. For example in the future an Expert System with learning capabilities could acquire all the knowledge of the best operators and process experts in any industry, so then if a substitute operator is on duty and encounters an abnormal situation that he or she is not familiar with, then this Expert System will be able to help with diagnosis and advice.

  • Customer research

With the introduction ANNs in customer research, companies have benefited tremendously. The applications are varied and have helped businesses to solve many problems. ANNs can be used to learn complex patterns of information about customers and generalize the learned information. With this information companies in the future will be able to predict customer satisfaction for instance. These systems will be able to identify the drivers of loyal customers; that is to say, what makes a customer become loyal. Using this information a system can be trained to analyze satisfied consumer behavior's and create customized advertizing that will turn these consumers into loyal customers.

  • Target marketing

The methods used for advertising can decide the future of many companies, because ANNs do not need a lot of information to identify patterns, it can be used to avoid un-target advertising. ANNs are currently being used to create market segmentation. Classification and recognition problems have been identified as the most commonly cited applications of artificial neural networks; other applications go as far as predicting bankruptcy and loan default and also modeling consumer choice and advertising responses. If the current recession in the US was caused in part by the housing market collapse, an ANN system would have been able to predict all those bad mortgages or potential customers that would default their loans, sounds unrealistic, but it might have been possible to prevent it from happening.

  • Risk management

In risk management, budgeting takes place in the early phases of a project, accurate estimates are very essential to ensure that a project gets sufficient funding; therefore, many project managers are seeking better ways to increase the accuracy of their project cost estimates. With the help of ANNs, they are able to predict how much a project will cost in the early phases of project design and planning. For instance, using ANN models project managers are able to identify at these early stages potential cost variations that might occur by the end of the project. Given these predictions they can put aside sufficient funds so that the project gets completed on time and also prevent over-budgeting and therefore devoting any extra funds to new projects/investments.


Wednesday, August 4, 2010

SB 1070

Is this the solution?


Arizona has passed the toughest and strictest anti-illegal immigration measure in decades in the United States, Arizona Senate Bill 1070, by doing so, controversy has erupted among those that support or condemn it. Even though, many claim that this bill do not promote racial profiling (mainly against Latinos), the core purpose of this law is deport as many illegal immigrant as possible. If the law would have been fully implemented, the police would need to consider three key provisions before calling the INS (they deport illegal immigrants). First, there has to have been a "lawful stop, detention, or arrest" to enforce another law. Second, during the course of that stop, detention, or arrest, a "reasonable suspicion" has to exist that "the person is an alien." And third, law enforcement "may not consider race, color, or national origin…except to the extent permitted by the United States or Arizona Constitution." Now, the question is, are these reasonable standards? The answer depends on who is responding. How many times have you (Any race other than Caucasian) been stopped (while driving) for no reason only to find out that apparently you crossed the double yellow line more than once? Or, how many times have you gotten away with a traffic violation because the person sitting next to you looked like the police officer?

So, instead of fighting over the legality or reasonability of these laws, let's look at the benefits and costs of legal and illegal immigration in the United States.

Anyone that has taken an introductory class in economics knows that in a cost benefit analysis, if the benefit is greater than the cost, it is a good investment/strategy from an economic point of view. Let's look at the argument of deporting all illegal immigrants. Many people seem to think that illegal immigrants (undocumented workers) should be simply deported to wherever they came from. This is not a good idea, both for our government and for our economy. To deport the approximately 14 million illegal immigrants, even if we knew where they all were, would cost billions of dollars in buses, airplanes, gas and food. Where would that money come from? Taxpayers? No one would like to pay more taxes, not even those that support massive deportations.

Besides, what harm are these immigrants doing us? Yes, they broke the law, a law that is currently broken, however, after arriving to this country they certainly contribute to our economy - after all, they buy things, start businesses, rent houses (which is important in today's housing market). YES! THEY ALSO PAY TAXES INCLUDING INCOME TAXES, there is a tax for almost anything that you buy. Yes, some illegal immigrants commit crimes. But so do legal immigrants, as well as the descendents of the Pilgrims. I fully support deporting illegal immigrants who are rapists, murderers and the like. However, most illegal immigrants are just ordinary people who seek to better their lives and those of their children, which is certainly the American Dream.

Consider an America without illegal immigrants. Let's examine our farms first. Who will be picking the crops? Americans used to working in offices, or even as janitors or McDonalds workers, are not going to be willing to work 12 hour days in the Florida heat picking oranges for minimum wage in substandard housing. To get enough workers, farmers must be willing to pay two, three, or many more times the wages they pay now. Farm products, therefore, will cost two, three or many more times what they cost now. Are you willing to pay $10 for a bunch of bananas during banana season? Even if you were, enterprising merchants would simply import foreign bananas and undercut the prices of American farmers. To deport illegal immigrants, therefore, would be to kill the American family farm. That surely isn't your intention, is it?

The same thing applies in cities. Would Mr. Merrill and Mr. Lynch of Merrill Lynch be willing to vacuum their own office? I doubt it. Mow their own lawns? Nope. While they would be able to pay for it, commodity prices would rise for you and I. We'd end up with hundreds of unclean offices, houses and unmowed lawns. Is that worth having a few fewer Spanish-speakers in America? Are you willing to change your standards of living because someone says "yellow Mr Yonson"?

These immigrants will integrate into American society in a few generations anyways and become as tightly woven into our societal fabric as the German and Italian immigrants of the early 1900s were, just as the Irish of 1860s before them were. Just like today, xenophobes claimed that the members of whichever immigrant group was demonized that week would never integrate into America, that they were destroying our American way of life, and that they would never learn English. Each time, however, the second-generation was bilingual, speaking English outside of the home, and Italian, German, Polish, Yiddish, Russian, Gaelic or some other language in the home, while their kids spoke English and only a few words of the language of the "home country."

The benefits of keeping immigrants coming in from whichever country keeps America a prosperous nation with one of the greatest standards of living in the world.

You be the judge, this is just my wee opinion.