"if a person has control over any function, it can also be used to control the computer"

The first thing that came to my mind when I saw this word was AI.
So I decided to look into AI.
AI is an abbreviation for Artificial Intelligence. At present, the definition of AI has not been set even among experts.

1. Types of AI

There are two types of AI.
Narrow AI
Artificial General Intelligence(AGI)

Narrow AI

Narrow AI refers to artificial intelligence that specializes in one thing. For example, technologies such as image recognition and voice recognition, autonomous driving technology, and shogi AI such as PONANZA can also be said to be Narrow AI. Currently, most of the research related to AI is dedicated to this Narrow AI.

Artificial General Intelligence(AGI)

Simply put, AGI is an AI that can do anything. Narrow AI can do only one thing, but AGI refers to AI that can think and apply based on given information. They behave like people themselves. Singularity is said to occur when AGI is completed, and it is said that AGI will be the last invention of humans.

Singularity

It refers to the time when the AI, which has intelligence beyond humans, was invented. It can be considered that when it is developed, this AI can produce even smarter artificial intelligence. In other words, it is said that AI with explosive (recursive) intelligence is developed, and AI that humans can never imagine will be created and life will change completely. Futurist Ray Kurzweil expects singularity to reach 2045.

2. History of AI

Currently, things and services that utilize AI are becoming popular in various industrial fields. Therefore, I would like to understand AI from the history.

First AI boom (1950s-1960s)

In this era, research was being conducted to solve problems by "inference / search" using computers. The word "artificial intelligence" started at a workshop held in Dartmouth in the summer of 1956. Prominent scholars such as John McCarthy, Marvin Minsky, Allen Newell and Harvard Simon were present. At this workshop, Newell and Simon demonstrated an artificial intelligence program called "Logic Theorist."  This is a program that automatically proves the theorem and was said to be the world's first artificial intelligence program. In the first AI boom, various algorithms in computer programs were devised, especially, it was good at escaping from mazes and solving puzzles called toy problems. However, it couldn't solve a real-life problem because there was no way to handle the complicated calculations.

Second AI boom (1980s)

The development and introduction of the "expert system" triggered the second AI boom. The developed "Expert System" is a program that answers questions about a specific area by using rules derived from expert knowledge, which was created with emphasis on knowledge representation. But there was a problem. The problem is that computers have no "common sense." For example, when someone asked, "To reduce fever," it answered, "Give antipyretic drug" or "Kill." Computers don't have that "common sense," so they answer this question. Eventually, facing such problems, the second boom has settled. 

Third AI boom (current)

The major causes of the third boom are the development of deep learning technology and the spread of big data. With this technology, it is possible to extract information from images and videos, and to generate music and characters. The aim of AI is to "learn and guess" by itself. It can be said that deep learning is a technology that makes it possible. There is no doubt that deep learning will play a major role in the future development of AI.

3. AI and work

It is even said that "artificial intelligence will take away work" due to the influence of evolving artificial intelligence technology. In fact, the future of AI taking jobs may be just around the corner. In a paper, Associate Professor Osborne of Oxford University in the UK has published a ranking of jobs that are likely to be lost to AI. The first place is a salesperson. If e-commerce is even easier to use, store salespersons will no longer be needed. Even senior officials such as central government officials who are said to be stable are ranked in. It is because if the machine can manage personal information etc. reliably, leakage can be avoided. It is no longer necessary for humans to carry out tasks that repeat one task. However, it is also true that a lot of complicated work has come along with it. Humans will have to take on the role of doing complex work from now on, so further skill improvement will be required.

4. AI and Programming

Python, which is popular for the development of AI, is a general-purpose programming language. The feature is that the code is simple and easy to handle compared to other programming languages. In the recent AI boom, we often hear the word machine learning, and Python is often used for machine learning. So why is Python often used in the AI ​​development field? Famous programming languages ​​include C, Java, etc. Python, on the other hand, has very few types, allowing you to write large numbers of programs concisely. Therefore, it is useful for AI development that needs to process huge amounts of data.


source


コメント

このブログの人気の投稿

Remember the human

Ubuntu And Kona Linux