Daniel Bollinger
11/26/2014
ENGL-2116-008-Fall
Ethics of AI (Artificial Intelligence) The first desktop computers were invented in 1977. Computers have been invented within one hundred years of today, and have become a part of our lives and daily routines. Even in the last decade computers have gotten smaller and more convenient. They can store more memory, have the ability to do more tasks, all taking less time than the previous version. Technology has been ingrained in American culture. The growth of technology has been rising exponentially. Revolutions in technology has proven itself to be valuable to society, so scientists and inventors have spent a lot of time, money and energy on artificial intelligence. Creating advanced artificial …show more content…
In psychology, this involves holding the relevant factors constant while varying other factors across experimental trials. For instance, when a person reacts to a visual display, many factors are at play, including how much caffeine he has consumed, the quality of his eyesight, and the lighting conditions. Experimenters keep relevant visual factors constant across trials and subjects in hopes that the irrelevant factors will cancel each other in the …show more content…
It is important that these questions be asked of a single integrated system rather than a collection of systems. It is generally much easier to tailor a system to deal with a narrow class of problems. However, these optimizations often come at the cost of failure on other aspects of a problem. For example, the classification abilities of a typical statistical neural network come at the cost of the reasoning abilities of a typical search algorithm and vice versa. One of the main challenges of HLAI research thus is to somehow overcome these tradeoffs and create a single system that can manifest all the aspects of intelligence. It is thus important that work in microcosms be evaluated, not according to the success of a class of systems on a set of problems, but instead according to the performance of a single system that integrates the ability to deal with each of these
11/26/2014
ENGL-2116-008-Fall
Ethics of AI (Artificial Intelligence) The first desktop computers were invented in 1977. Computers have been invented within one hundred years of today, and have become a part of our lives and daily routines. Even in the last decade computers have gotten smaller and more convenient. They can store more memory, have the ability to do more tasks, all taking less time than the previous version. Technology has been ingrained in American culture. The growth of technology has been rising exponentially. Revolutions in technology has proven itself to be valuable to society, so scientists and inventors have spent a lot of time, money and energy on artificial intelligence. Creating advanced artificial …show more content…
In psychology, this involves holding the relevant factors constant while varying other factors across experimental trials. For instance, when a person reacts to a visual display, many factors are at play, including how much caffeine he has consumed, the quality of his eyesight, and the lighting conditions. Experimenters keep relevant visual factors constant across trials and subjects in hopes that the irrelevant factors will cancel each other in the …show more content…
It is important that these questions be asked of a single integrated system rather than a collection of systems. It is generally much easier to tailor a system to deal with a narrow class of problems. However, these optimizations often come at the cost of failure on other aspects of a problem. For example, the classification abilities of a typical statistical neural network come at the cost of the reasoning abilities of a typical search algorithm and vice versa. One of the main challenges of HLAI research thus is to somehow overcome these tradeoffs and create a single system that can manifest all the aspects of intelligence. It is thus important that work in microcosms be evaluated, not according to the success of a class of systems on a set of problems, but instead according to the performance of a single system that integrates the ability to deal with each of these