Stages and Categorizations of AI
- [Stages of AI - PwC]
Don't be afraid of fail. Be afraid not to try.
- Overview
The modern project to create human-like artificial intelligence (AI) began after World War II, when it was discovered that electronic computers were not only number-crunching machines, but could also manipulate symbols. This can also be achieved without assuming that machine intelligence is the same as human intelligence.
Artificial intelligence (AI) has been named the most widely mentioned technology in recent times, according to a recent study by technology analyst Gartner. Most CIOs agree that AI has the greatest paradigm-shifting power. According to most predictions, AI should occupy the center stage of most human endeavors in the next few decades.
But AI is far from a static technology with a fixed set of principles. In addition to providing the core value of mimicking human intelligence and reasoning to get work done faster, faster, and better, AI continues to evolve over time, becoming more capable and richer. This is called weak AI.
However, the goal pursued by many AI researchers is to develop AI that is in principle the same as human intelligence, called strong AI. Weak AIs are less ambitious than strong AIs and therefore less controversial. However, there are also important controversies associated with weak AI.
- Categorizations of AI
Artificial intelligence (AI) technology creates opportunities to solve real-world problems in health, education, and the environment. In some cases, AI can do things more efficiently and methodically than human intelligence.
“Smart” buildings, vehicles and other technologies can reduce carbon emissions and support people with disabilities. Machine learning is a subset of AI that enables engineers to build robots and self-driving cars, recognize speech and images, and predict market trends.
AI is classified by many norms. Two of the main categorizations of AI are capability and functionality.
AI has three types on the criteria of capability: Artificial Narrow Intelligence (Narrow AI), Artificial General Intelligence (General AI), and Artificial Super Intelligence (Super AI).
There are four types of AI on the basis of functionality: Reactive Machines, Limited Memory, Theory of Mind, and Self-awareness.
- Three Types (Stages) of AI - Based on Capabilities
There are various ways to create AI, depending on what we want to achieve with it and how we will measure its success. It ranges from extremely rare and complex systems, such as self-driving cars and robotics, to parts of our everyday lives, such as facial recognition, machine translation, and email categorization. The path you choose will depend on what your AI goals are and how well you understand the intricacies and feasibility of various approaches.
AI technologies are categorized according to their ability to mimic human traits, the techniques they use to do so, their real-world applications, and theory of mind. Using these characteristics as a reference, all AI systems — real and hypothetical — fall into one of three categories:
- Narrow artificial intelligence (ANI), with a narrow range of capabilities;
- Artificial General Intelligence (AGI) comparable to human capabilities; or
- Artificial Superintelligence (ASI), more capable than humans.
Today, we have three different variants of AI technology; ANI, AGI, and ASI. These are the three stages in which AI can evolve. We have only achieved narrow AI so far.
As machine learning capabilities continue to develop and scientists move closer to achieving AGI. Theories and speculation about the future of AI are circulating. ASI is a futuristic idea about the ability of artificial intelligence to replace human intelligence. For ASI to become a reality, computational programs must surpass human intelligence in all parameters and environments.
- Four Types of AI - Based on Functionalities
The four primary types of artificial intelligence (AI) based on functionality are reactive machines, limited memory, theory of mind, and self-awareness:
- Reactive machines: The most basic type of AI, reactive machines can respond to immediate requests and tasks, but they can't store memories or learn from past experiences. Reactive machines are AI systems that have no memory and are task-specific, meaning that the input always delivers the same output. Machine learning (ML) models tend to be reactive machines in that they take customer data (such as purchase or search history) and use it to provide recommendations to the same customer. This type of artificial intelligence is reactive. It performs "super" AI because ordinary people cannot handle large amounts of data. Reactive AI is largely reliable and works well in inventions like self-driving cars. It cannot predict future results unless appropriate information is provided.
- Limited memory: The next type of AI development is limited memory. Unlike reactive machines, limited memory AI systems can use short-term or recent data to make decisions. They have a temporary memory that allows them to adapt their responses based on recent information, like a self-driving car adjusting its actions based on recent traffic data. The algorithm mimics the way the neurons in our brains work together, meaning it gets smarter as it is trained on more data. Deep learning algorithms improve natural language processing (NLP), image recognition, and other types of reinforcement learning. Unlike reactive machines, AI with limited memory can look back and monitor specific objects or situations over time. These observations are then programmed into the AI so that it can perform actions based on past and present data. But in the limited memory, this data will not be saved in the memory of AI as an experience for learning, the same way humans derive meaning from success and failure. Over time, AI improves as it is trained on more data.
- Theory of Mind: The first two types of AI, reactive machines and limited memory, are the types that currently exist. Theory of mind and self-aware AI are the types of theories that can be built in the future. Therefore, there aren't any real-world examples yet. If developed, AI's theory of mind has the potential to understand the world and how other entities have thoughts and emotions. This, in turn, affects their behavior in relation to those around them. Humans have the cognitive ability to process how our own thoughts and emotions affect others, and how the thoughts and emotions of others affect us—the foundation of our social relationships. In the future, theory-of-mind AI machines will be able to understand intentions and predict behavior, much like simulating human relationships.
- Self-aware: Considered the most advanced level of AI, self-aware AI can create machines that deeply understand and are aware of complex human emotions and mental states, including their own. This type of AI would be able to understand its own emotions and intentions, and interact with humans in a very natural way. The ultimate outcome of the evolution of AI will be the design of systems that are self-aware and have a conscious understanding of their existence. This type of AI does not yet exist. This goes beyond AI theory of mind and understanding emotions to allow us to be aware of ourselves, our own state of being, and be able to sense or predict how others are feeling. AI and machine learning algorithms are still a long way from self-awareness There is a long way to go, because there is still much to be revealed about the intelligence of the human brain and how memory, learning, and decision-making work.
These four types of AI together enable technologies such as Natural Language Processing (NLP), computer vision, facial recognition, machine learning, and deep learning.
- [Rome, Italy - Civil Engineering Discoveries]
- Artificial Narrow Intelligence (ANI)
In contrast to strong AI, which can learn to perform any task humans do, weak AI (or narrow AI) is limited to one or a few specific tasks. This is the kind of artificial intelligence we currently have. In fact, deep learning, named after the human brain (and often compared to it), has very limited capabilities and is nowhere near what a human child's brain can do. This is not a bad thing.
In fact, narrow AI can focus on specific tasks and do it better than humans. For example, feed a deep learning algorithm enough pictures of skin cancer and it will be better at spotting skin cancer than an experienced doctor. This does not mean that deep learning will replace doctors. You need intuition, abstract thinking, and more skills to decide what is best for your patients. But deep learning algorithms are sure to help doctors do their jobs better and faster, and care for more patients in less time. It will also reduce the time required to educate and train healthcare industry professionals.
- Artificial General Intelligence (AGI)
Artificial general intelligence (AGI), also known as strong AI, is a theoretical form of AI that would allow machines to perform tasks with human-like cognitive abilities. AGI machines would be able to learn, solve problems, plan for the future, and have self-awareness and consciousness. They would also be able to mimic human intelligence and behave in similar ways to humans.
Some say that AGI is still decades or even centuries away from being achieved. However, others believe that AGI will become a reality in our lifetime. Some capabilities that AI may need to master before achieving AGI include: creativity, algorithmic advances, and new robotics approaches.
General AI (AGI) is only theoretical at this point. This is the AI that writers have made up for years in sci-fi stories. Ultimately, when we achieve AGI, machines will have consciousness and decision-making capabilities - full human cognitive abilities. These machines do not require human input to be programmed to function.
For all intents and purposes, this will be a time when machines act, feel, respond and think like humans. We can say that a powerful AI has a mind of its own and is capable of doing whatever it wants to do like any human being. Unlike narrow AI, which classifies data and finds patterns, general AI uses clustering and association when processing data. AGI will also be self-aware. However, like a child, AI must learn through experience, improving knowledge and skills over time.
But while all of this talent is focused on finding a way to create a powerful AI that can compete with the human brain, we are missing a lot of opportunities and failing to address the threats posed by current weak AI technologies.
Some commentators have argued that weak AI could become dangerous because of this "fragility" and fail in unpredictable ways. Weak AI could disrupt power grids, damage nuclear power plants, cause global economic problems, and mislead self-driving cars. In 2010, a weak AI trading algorithm caused a "flash crash" that caused a temporary but significant drop in the market.
- Artificial Super Intelligence (ASI)
ASI is a futuristic concept and idea of AI replacing human intelligence capabilities. For ASI to become a reality, computational programs must surpass human intelligence in all parameters and environments. ASI will only become a reality when AI becomes smarter than humans.
ASI with a futuristic halo seems far removed from the future of human evolution, in the sense that this so-called perceptual AI variant shows the conceptual limits of AI technology and its promises that it won't deliver, at least in become a reality decades later.
If ASI becomes possible and becomes a reality, the role of humans in decision-making, the arts and humanities, and emotional understanding of all aspects of life could be at odds with the rise of machines.
