The major goals of artificial intelligence also correspond to the traditional problems AI research intends to solve. Each goal corresponds to a specific subset or subfield of artificial intelligence. These core areas include knowledge representation and reasoning, machine learning, natural language processing, computer vision, artificial general intelligence, and artificial intelligence safety. A foundational understanding of artificial intelligence necessitates a grasp of these diverse goals and subfields.
Understanding the Major Goals of Artificial Intelligence: Main AI Research Problems and Subfields of AI
1. Knowledge Representation and Reasoning
One of the main goals of artificial intelligence is to design computer systems and models that can use their knowledge of the world to process information and solve complex problems. This is known as knowledge representation and reasoning.
Furthermore, as a specific AI subfield, it is involved with the automation of different kinds of reasoning processes through the codification of factors or relationships between ideas in a manner that can be interpreted by a computer system.
Some applications of knowledge representation and reasoning include a computer-aided diagnosis for assisting physicians and natural language user interfaces that enable using human language as input for interacting with computers or software.
2. Automated Planning and Scheduling
Planning and scheduling is another subfield of AI focused on the automated generation of action sequences and formulation of strategies that can be executed by an artificial intelligence system. This is one of the traditional problems of AI research.
AI planning aims to automate the generation of a plan based on predetermined goals and a set of possible actions. Take note that this is also one of the fundamental capabilities needed to increase the autonomy and flexibility of advanced AI systems or AI agents.
Specific examples of AI planning include self-correcting software applications, robots that function as autonomous agents, autonomous or self-driving vehicles, automated information gathering systems, and computer-aided suggestion or recommendation systems.
3. Machine Learning and Deep Learning
Machine learning is both an AI subfield and a practical AI application. This subfield focuses on the study of algorithms and models that computer systems can use to perform specific tasks without explicit programming or human instruction.
At the heart of machine learning is a different approach to computer programming. The specific application centers on developing algorithms that can process and analyze Big Data and learn from the outcomes without explicit programming.
Deep learning is an advanced subset of machine learning that uses artificial neural networks to process and analyze data. Examples of machine learning applications include search engine ranking, content or data generation, and autonomous driving
4. Natural Language Processing
Another goal and subfield of artificial intelligence is natural language processing or NLP. It focuses on human-computer interaction and even computer-to-computer interaction using natural human language instead of a computer language.
The processing of natural language is a traditional problem in AI. The NLP subfield aims to develop systems that can process large amounts of natural language data and language models that can understand and generate natural human language.
Applications of NLP include speech recognition and language translation. Further advancements have led to the development of large language and multimodal language models used in various generative artificial intelligence applications like chatbots.
4. Computer Vision
The need to equip a computer with capabilities to understand or process and analyze visual data from still or moving live and digital images is another one of the main goals of artificial intelligence. The subfield of computer vision aims to accomplish this.
Achieving computer vision requires developing algorithms that can mimic the processes involved in natural vision and models trained from a large set of visual data. This then equips a particular system with capabilities to process, interpret, and utilize visual data.
One of the notable applications of this AI subfield is facial recognition. Other applications include automated image manipulation, video tracking, autonomous driving, and integration with virtual reality and augmented reality or mixed reality systems.
5. Robotics
Another important goal of artificial intelligence is robotics. This AI subfield draws from various branches and disciplines of science and engineering. These include computer science, mechanical engineering, physics, and electronics engineering.
Robotics involves the design, construction, and operation of machines that can replicate human movements to replace human tasks with mechanical ones. AI equips these machines with autonomous operation, sensory awareness, and decision-making capabilities.
Current research in the subfield aims to introduce commercial, domestic, and military applications. Amazon has been using autonomous robots in its warehouse facilities. Large manufacturers have also integrated robots into their assembly lines.
6. Artificial General Intelligence
The utmost and long-term goal of artificial intelligence as a field is superintelligence or artificial general intelligence or AGI. A superintelligent system can perform intellectual tasks that are comparable to or even better than the intellectual capabilities of humans.
An artificial general intelligence system can either be self-conscious, self-aware, or both. This system will demonstrate the capabilities of machine-to-human and machine-to-machine interactions that replicate normal human-to-human interactions.
Furthermore, based on the three major types of artificial intelligence systems, artificial general intelligence exemplifies the development of a humanized system equipped with cognitive intelligence, emotional intelligence, and social intelligence.
7. AI Safety and AI Ethics
The prospect of developing computers or machines that can think and the rapid advancements in artificial intelligence have led to the emergence of another AI goal and the creation of additional AI subfields. These are AI Safety and AI Ethics.
AI safety has emerged as an interdisciplinary subfield of artificial intelligence that focuses on preventing accidents, misuse, or other harmful consequences that could arise from the development and deployment of artificial intelligence systems.
There are more specific subsets within this AI subfield. Machine ethics is concerned with embedding systems with human-based moral and ethical standards, while AI alignment focuses on aligning systems with human values and human preferences.
Overlaps Across the Different Goals and Subfields of Artificial Intelligence: Specific Overlaps in Advanced AI Systems
Advancements in the field of artificial intelligence have resulted to the development and deployment of systems and models that demonstrate the use of various concepts and even specific objectives unique to each AI goal and AI subfield.
Notable generative AI applications, such as AI chatbots, have required the use of deep learning trained from large natural language data and the development of large language models for natural language processing. Multimodal large language models are also trained with visual data to equip them with computer vision capabilities.
The subfield of robotics has also showcased advanced robots trained on large datasets using deep learning techniques. Some of these robots are also equipped with computer vision capabilities using sensors for environmental awareness.
An autonomous vehicle demonstrates the utilization of various tools and techniques from the different subfields of artificial intelligence. For starters, it has sensors for object tracking, calculating proximities, and object recognition. It also has a user interface that uses natural language processing for human-machine interaction.
The arrival of AGI will further demonstrate various overlaps among different AI goals and subfields. A particular AGI system can possess natural language processing capabilities, native deep learning functions, and advanced computer vision.