Types of AI classified by role
There are two types of AI classified by role: specialized AI and general-purpose AI.The role here refers to the scope of the processing work.
For example, the roles of people who work in a supermarket can be divided into those who work at the cash register, those who purchase goods, and those who display products.
Specialized AI refers to those that have a fixed role, that is, the scope of work that can be processed is limited.Specialized AI has its own role, such as predicting the weather from weather data or determining the type of animal in a picture.
All AI currently in practical use is categorized as specialized AI. In the case of current AI, it is necessary to clearly define the purpose of processing when designing, and then let the AI learn.
General-purpose AI refers to the type of artificial intelligence that can play any role, or handle any task.Since you can perform a variety of roles, you can understand the situation in which you are placed under any circumstances, and you can think and act on your own.
General-purpose AI has not been put into practical use. If general-purpose AI is put into practical use, it is said that the AI will be able to learn on its own, think about its purpose, and perform processing without being taught anything.
What is the difference between strong AI and weak AI?
Apart from general-purpose AI and specialized AI, there is a classification method of strong AI and weak AI. This idea was put forward by philosopher John Searle in 1980.The difference between strong and weak AI is whether it is self-conscious, or has a mind.
By definition, strong AI and general purpose AI, weak AI and specialized AI are the same thing. Therefore, AI as strong as general-purpose AI has not yet been realized.
A strong AI is an AI that is self-conscious.Examples include Doraemon and robots in science fiction movies. Also, strong AI and general-purpose AI can be considered the same thing.
Strong AI is not currently available. In order to be as self-conscious as humans are, we first need to clarify human intelligence.
Since current science has not clarified the mechanism of intelligence, it is not possible to equip AI with a mechanism to create strong AI.
Weak AI is AI that has no self-awareness.It is created by designing by narrowing down the purpose of the work to be processed by the AI designer. All of the AIs in use today are weak AIs.
Weak AIs have no real self-awareness and cannot decide to learn or start from scratch.
It can only learn and process within the range designed by AI designers. In other words, specialized AI and weak AI are the same thing.
9 main uses of AI
There are nine main applications of AI: games, chatbots, stock price prediction, speech recognition/translation, recommendation functions, image recognition, image captions, automatic coloring, and object detection.
So far, only specialized AI has been realized, but AI is already being used in many fields and is attracting attention.
If the application is suitable for AI, some of them are showing performance similar to or exceeding that of humans.
Application of AI 1: Games
Games are a typical example of where AI can be used.Games have a clear purpose, and the content to be learned can be limited to the rules of the game, so it can be said that it is a suitable application for AI.
Gaming AI has already surpassed humans in many types of games. AI has beaten human opponents in the popular board games of chess, shogi, and go.
Application of AI 2: Chatbots
AI is also used in chatbots that use text and voice to converse.By equipping chatbots with AI, they can naturally respond to a wider range of questions.
Specifically, it is used as an inquiry function on corporate sites, and for telephone support at call centers.
A chatbot AI called Watson developed by IBM became a hot topic around 2015 when it was introduced to a Japanese megabank.
Application of AI 3: Stock price prediction
AI is also used for stock price prediction.Stock prices fluctuate under the influence of various factors. You can predict stock prices by letting AI learn using stock-related information and images of stock price charts.
Since the judgment criteria are not ambiguous like humans, it is expected to leave stable results.
There are also securities companies that manage stocks by incorporating stock price predictions using AI. It seems that there is also a service that allows you to receive advice from stock price prediction AI.
Application of AI 4: Speech recognition/translation
The kind of AI that recognizes voice and translates spoken words is also being used.Sometimes it shows sentences that make you feel uncomfortable, but the translation is at a level that humans can generally understand.
Google Translate is free to use and supports Japanese, English, and many other types of languages.
There is also a function that automatically translates overseas websites and websites, making it easier to obtain information in Japanese even if it is overseas.
Application of AI 5: Recommendation function
AI is also used in the recommendation function, which is a function that selects recommended products for customers.It displays recommended products based on the behavior of buying or browsing a certain product.
It is used to increase the sales volume of products by presenting recommended products to customers.
When you shop on Amazon, the recommendation function displays a list of recommended products, such as “People who bought this item also bought this item.”
AI Application 6: Image Recognition
By using the kind of AI that can recognize images, you can know the type of animal in the photo, and you can convert handwritten characters into text data.
Image recognition is the ability to determine what is in an image.
A smartphone app called Google Lens allows you to search for objects similar to what you have taken using the image recognition function. In addition, the function that the camera application automatically recognizes the QR code has also been put into practical use.
AI Application 7: Image Captioning
A function called image caption has been realized, in which AI automatically generates a descriptive text for the given image.
In addition to image recognition, we need the ability to determine the situation in a given image. The specific content of the caption can be the gender of the person in the photo or the action the person in the photo is doing.
Show a picture of a running zebra and automatically add a description such as ‘the zebra is running’.
Application of AI 8: Automatic coloring
By using AI, you can automatically add color to images that do not have color.It is a function called automatic coloring, and specifically, there is a function that automatically colors line drawings and monochrome images.
Until now, to convert black-and-white video to color video, it was necessary to manually color each frame, but by using automatic coloring AI, the time can be greatly reduced.
This makes it possible to enjoy old images more easily as realistic images.
AI application 9: Object detection
AI that applies the image recognition function to detect the type and position of an object in the image has also been put to practical use.It is a function called object detection, and it has been put to practical use in the field of manufacturing and medical care.
In the case of the manufacturing industry, if AI for object detection is used, there is no need for inspectors to check products one by one during the inspection process. increase.
In the medical field, abnormalities in the body can be detected from images to reduce the risk of being overlooked.
What is an AI assistant?
An AI assistant is a function that gives instructions and asks questions by voice and performs tasks instead of humans.
There are those that function independently of devices equipped with AI assistants, and those that function in conjunction with other electronic devices.
AI assistant modern technology
Modern AI assistant technology is realized by combining several types of AI, such as voice recognition, chatbot functions, and search functions.
It recognizes the user’s voice and verbalizes it, searches for information if necessary, and performs appropriate actions.
4 types of AI assistants
There are four types of AI assistants currently in practical use: apps, smart speakers, in-vehicle AI assistants, and robots.
Both kinds of AI assistants are used in a way that saves the individual from having to do something and do difficult operations. From here, we will introduce four types of AI assistants that are currently in practical use.
AI Assistant Type 1: Apps
A smartphone app with an AI assistant function has been put into practical use.Popular examples include Siri on iPhone and Google Assistant on Android. It recognizes your voice, makes it easy to search, and allows you to operate your smartphone a little.
The function to enter text with voice recognition AI instead of entering text is also convenient. You can’t type as fast on your smartphone as you can on a computer, but you can use voice recognition to quickly type characters.
AI Assistant Type 2: Smart Speaker
Smart speakers equipped with AI are convenient when you want to easily operate home appliances in your home or when you want to easily check the weather forecast.
The basic functions are the same as Siri and Google Assistant, but it is characterized by being able to operate in conjunction with electronic devices other than the smart speaker itself.
By linking with lighting equipment, you can operate the lights without a remote control, and you can also link with your smartphone and make a call just by speaking.
AI assistant type 3: In-vehicle AI assistant
An in-vehicle AI assistant is an AI that provides voice assistance while driving.While driving, it is dangerous to let go of the steering wheel or shift your eyes to the navigation panel.
The function that helps you operate with only voice is not only convenient, but also enhances the safety of driving. You can operate the car navigation system hands-free, play music, and perform minor operations on your smartphone.
AI Assistant Type 4: Robot
Of course, robots are also equipped with AI.Pepper released by Softbank is probably the most famous.
Some companies are installing robots at the front desk of hotels and companies to guide customers. There are also examples of using AI-equipped robots as serving staff at restaurants and other establishments.
Let’s grasp the types of AI that are used in various places!
In this article, we have introduced the types and applications of AI.AI can be divided into general-purpose AI, specialized AI, or strong AI or weak AI, and it turns out that all AI in practical use is specialized AI.
AI is already being used in various fields, and AI assistants that can be used in ordinary homes are also appearing.
By understanding the types of AI, you will be able to get a more concrete image of how to effectively use AI in your life and work.