Can We Design an Artificial Intelligent Program to Find Experts by Their Writing?
f we are to build an artificial intelligent computer program to recognize experts by the words that they write there are a few things we must do. First, we must consider how a human being comes to the conclusion if someone is an expert and determine if we can apply these same procedures. How does one determine by reading someone’s work if they are an expert or a poser?
Most humans can readily tell if someone is an expert or not by reading their articles, summaries, books, blogs, emails or research papers. What is it that triggers an individual into that moment when the light-bulb goes off and the human says, Wow, this person really knows their stuff?
Just because someone knows their stuff, do we consider them an expert? If this is the case then we could say that the encyclopedia is an expert? Is it? No, but it could be considered an authority on a subject, we call such; Reference Books. How can we therefore tell if someone is an expert or merely a book-smart, rote memorizing authority?
It turns out that most humans can tell when someone is an actual expert and not just an authority on the subject, but that takes a little more time than to say that they are or are not knowledgeable in a subject matter. Can we therefore design a computer that can do all three judging someone’s written words?
- Writer is knowledgeable in subject matter Yes or No
- Writer is an Authority Yes or No
- Writer is an Expert Yes or No
You see with computers we must do this sequentially. First the person has to be knowledgeable in the subject matter, if not, they do not qualify. If they are knowledgeable then we need to know if they are an authority. Once we determine that they are an authority, we can then determine if they are an expert.
Why is this important? Well, because in any given field there are lots of people contributing to that industry. Some will tell us of the 80/20 rule, where 20% of the people are really the good at what they do. But being good at what they do not necessarily mean that they are an expert. Many people realize that the top dogs in any industry are only about 5%.
So, how do we design a computer to find the top 5% and how can we be sure that folks are not playing the system or using deception to make the grade? Well in my estimation it might be smart to simply not tell anyone you are grading them, therefore they will not work to game your artificial intelligent system. After all a deceptive human can lead another human to believe they are an expert, when in fact they are not.
Currently, there are artificially intelligent resume readers which use key words in resumes to pick out what they are looking for and thus they can go online and read resumes or sort through huge stacks of resumes by scanning them in and reading the words and picking out the words by searching for specific words. Unfortunately in reading what someone has written it is doubtful that they will have on the article that they are a 25 year veteran of the industry with a PhD right?
If you need more information visit our site: AI Code Writing