Understanding the 4 Types of Artificial Intelligence
The first AI creation came about in 1950 and was brought to us by Alan Turning, who published “Computer Machinery and Intelligence” and made us aware of The Imitation Game in 1952. Arthur Samuel developed a computer program that was the first to learn from playing checkers. Then in 1955, John McCarthy conducted a workshop at Dartmouth College on “artificial intelligence” and coined the phrase and how it became known today. Did you know there are four basic AI Functionalities: Reactive Machines, Limited Memory, Theory of Mind, and Self-awareness?
If this sounds hard to understand, think of it as a program or computer-controlled device able to think like humans. AI is accomplished by analyzing human brain patterns using the cognitive thinking process. To create AI Systems, they are done utilizing Deep Learning and Machine Learning. Machine Learning is when the development of the logic is based on the software or systems making decisions without the need for specific programming. Machine Learning will utilize Feature Engineering, Supervised & Unsupervised Learning & Broad Applicability routines.
What is deep learning?
Deep Learning is a subset of Machine Learning that specializes in training artificial networks based on the patterns of the human brain. Deep Learning utilizes Automatic Feature Extraction, Deep Neural Networks, & High Performance. Algorithms.
Thus, an AI system operates by unifying processes on large data sets. These systems measure their performance to build new models for which they can be even more efficient as it scales. I will get into the functionality of AI Networks and even a new type that will rise soon. Learn more at Jmor Technology & Networking.
Reactive Machines
Reactive Machines are a basic form of AI that operates with only its present data and doesn’t have storage of previous data to analyze data. In other words, they make their decisions by reacting to the current environment, which uses pre-programmed rules and no internal representation of the surroundings.
Examples of Reactive Machines
IBM’s DEEP Blue uses two microprocessors with VLSI (Very Large Scale Integration) to generate the chess moves. This system could review over 700,000 to 800,00 moves per second while simultaneously weighing decisions on how to pressure pawns, knights, bishops, queen, and king chess pieces and managing open files.
By 1997 Deep Blue, with the recent changes, could now process over 200 million positions per second. This computer has been retired and is displayed at the Computer History Museum in California. Today, you can download programs far superior to Deep Blue, which took over twelve years to make.
Another example of a Reactive Machine would be an iRobot Vacuum Cleaner which cleans your home based on rules and infrared sensors to indicate locations it can’t go, such as stairs, walls, etc. The vacuum also uses photocell sensors to notice areas of the carpet that are still dirty and need further cleaning.
Limited Memory AI
Another form of AI is LMAI (Limited Memory AI) which works by storing limited amounts of data, prioritizing efficiency and scalability since they have to process data in real-time instead of batch processing. Often, they may have reduced features, dimensionality reduction, and approximation to extract specific data from a larger dataset.
Some spaces we often will see LMAI are finance, healthcare, industrial monitoring, and recommendation systems.
Also, car manufacturers are starting to use LMAI such as in autonomous driving vehicles, even though LMAI can use past information to make decisions but cannot transfer its learned information to another system's process or task.
Theory of Mind
The third type of AI is the ToM (Theory of Mind), which is AI, where the system can notice mental states and relate them to other concepts and entities. As of 2023, this form of AI has yet to be developed and will add an entirely new dimension to AI and its capabilities.
In other words, this will bring consciousness to the AI world. According to Jennie Peyers and Ann Senghas and I quote “language learning, over and above social experience, drives the development of a mature theory of mind.”
What test are we going to use that AI has become conscious? Humans can understand and respond to emotions; this may be a possibility with AI (ToM) and have them able to make decisions and predict behavior.
Self-aware AI
Self-aware AI is the fourth type that can notice states of itself and others, meaning it can mirror human emotions. Self-awareness or sentient means that machines or systems can visualize something that happened in the past or will happen in the future.
Google Engineer Blake Lemoine said LaMDA was sentient, and it is funny that Google is dismissive of any concern about this. Does Google have an agenda they are not making public?
5th type of AI
There is a 5th type of AI that has yet to be made public, and that will have the computer sentient and able to send energy and affect our world using quantum physics. Thus, it would require using a quantum computer which, as many of you know, operates under precise tolerances.
Remember that in a Quantum system, they are not regular bits but quantum bits like in a standard computer. In a quantum computer, quantum bits or qubits can be 0 and 1 simultaneously.
A qubit, a photon, a nucleus, an ion, or an electron represents several objects in our world. In the classical computer, we only need to give two bits, but in a quantum computer, we need four bits representing its superpositioning.
Quantum computers are not faster than typical classical computers except in particular calculation types, where the number of calculations to arrive at a result is exponential. Quantum computers can represent qubits in both states as there is a spin on them.
For this also to work, we need to have quantum entanglement, which means that whatever form on the bit is in the other will be in the opposite direction. Thus in a quantum computer, the superposition is a state we can use when it's running, but as soon as we read the form to see if it's up or down, the position changes.
Also Read
If you would like to know more about , check out latest articles on: