In the field of computer science, artificial intelligence refers to the behavior of machines that can respond to their environment in a way similar to human intelligence and improve their probability of achieving goals as much as possible.
Weak Artificial Intelligence
Strong Artificial Intelligence
Super Artificial Intelligence
1. Weak Artificial Intelligence Weak Artificial Intelligence (Weak AI) is also known as Narrow Artificial Intelligence (Narrow AI) or Applied Artificial Intelligence (Applied AI), referring to AI that can only complete a specific task or solve a specific problem.
2. Strong Artificial Intelligence Strong Artificial Intelligence (Strong AI) is also known as Artificial General Intelligence (AGI) or full artificial intelligence, referring to intelligent machines that can perform any intellectual task as well as humans. This type of AI is a major research goal in the field of AI and has been a perennial topic in many science fiction works.
Thinking ability, the use of strategies to solve problems, and the ability to make judgments in uncertain situations;
A certain amount of knowledge;
Planning ability;
Learning ability;
Communication ability;
The ability to achieve goals using all its capabilities.
3. Super Artificial Intelligence Philosopher Nick Bostrom, director of the Future of Humanity Institute at Oxford University, defines superintelligence as “any intelligence that greatly exceeds human cognitive performance in almost all areas.” First, superintelligence can achieve functions equivalent to human intelligence, meaning it can reprogram and improve itself biologically, which is the “recursive self-improvement capability.”
4. A Brief Review of the Three Major Schools of Artificial Intelligence In reviewing the history of AI development, we find that it has primarily developed through the interplay of three aspects: Symbolism, Connectionism, and Behaviorism.
Aiming to express the formation of thought using logical symbols from mathematics and physics, generating human-like intelligence through a large number of “if-then” rules, this is a top-down process, including expert systems, knowledge engineering, etc.
Arguing that intelligence comes from the connections between neurons, it allows computers to simulate the neural networks and their connection mechanisms in the human brain, which is a bottom-up process, including artificial neural networks, etc.
Refers to control systems based on perceptual behavior, enabling each basic unit to achieve self-optimization and adaptation, which is also a bottom-up process, with typical representatives being evolutionary algorithms, multi-agent systems, etc. The artificial intelligence design model composed of these three aspects is shown in the figure.
Machine Learning and Deep Learning We previously introduced the definition and basic concepts of artificial intelligence, and now we will introduce the main branches of AI development: machine learning and deep learning.
5. Machine Learning and Deep Learning A very important branch of AI development is machine learning, an important field branch developed from the connectionism of AI, with the core purpose of enabling computers to have learning capabilities similar to humans. A large branch within machine learning is neural networks; strictly speaking, deep learning is a category of machine learning, but with the development of deep neural networks in recent years, especially the continuous expansion of deep learning applications, deep learning has become an important part of the field of machine learning.
Written by: Dai Weikun
