Types of artificial intelligence
Pro Research Analysisby 
Searched over 200M research papers
Analyzed relevant papers
Types of Artificial Intelligence: Key Categories and Characteristics
Artificial intelligence (AI) can be classified in several ways, based on its capabilities, autonomy, and the types of tasks it can perform. Understanding these types helps clarify how AI is used in different fields and what its future potential might be.
AI by Cognitive Capacity: Weak, General, and Superlative AI
AI systems are often grouped by their cognitive abilities:
- Weak or Limited AI: These systems are designed for specific tasks and cannot perform beyond their programmed functions. Examples include facial recognition, digital assistants, and text editors. They excel at narrow tasks but lack broader understanding or reasoning abilities 23.
- General AI: This type of AI would have the ability to understand, learn, and apply knowledge across a wide range of tasks, similar to human intelligence. While this remains a goal for researchers, true general AI does not yet exist 36.
- Superlative AI: Sometimes discussed as a future possibility, this would surpass human intelligence in all respects. It remains theoretical at this stage .
AI by Autonomy: Reactive, Deliberative, Cognitive, and Autonomous Systems
AI can also be distinguished by how independently it operates:
- Reactive AI: These systems respond to specific inputs with pre-defined outputs, without memory or learning from past experiences .
- Deliberative AI: These systems can plan and make decisions based on models of the world, showing more advanced reasoning .
- Cognitive AI: These systems can learn from data, adapt to new situations, and improve over time .
- Totally Autonomous AI: These systems can operate independently, making decisions and taking actions without human intervention .
AI by Task Intelligence: Mechanical, Analytical, Intuitive, and Empathetic
AI’s ability to perform tasks can be broken down into four types of intelligence:
- Mechanical Intelligence: Involves repetitive, rule-based tasks that are easiest for AI to automate, such as data entry or basic calculations .
- Analytical Intelligence: Involves analyzing data, recognizing patterns, and making logical decisions. Many current AI applications, like data analytics and recommendations, fall into this category 25.
- Intuitive Intelligence: Involves making judgments or decisions in uncertain situations, often based on experience or incomplete information. AI is beginning to develop these capabilities, but humans still outperform machines in this area .
- Empathetic Intelligence: Involves understanding and responding to human emotions. This is the most challenging for AI and is still largely a human domain, though research is ongoing .
AI by Methodology: Machine Learning, Deep Learning, and Natural Language Processing
AI technologies are also categorized by the methods they use:
- Machine Learning (ML): AI systems that learn from data to improve their performance over time. ML is widely used in healthcare, finance, and many other fields 247.
- Deep Learning: A subset of ML that uses neural networks with many layers to analyze complex data, such as images or speech. Deep learning powers applications like self-driving cars and advanced medical diagnostics 47.
- Natural Language Processing (NLP): Enables machines to understand and generate human language, used in chatbots, translation, and information extraction 247.
Theoretical Perspectives: Types 1 and 2 AI
Some researchers have proposed theoretical frameworks to classify AI:
- Type 1 and Type 2 Theories: These frameworks distinguish between different approaches to solving information processing problems, offering a rigorous perspective on AI’s development and future prospects .
Conclusion
AI can be classified by its cognitive capacity (weak, general, superlative), level of autonomy (reactive to fully autonomous), type of task intelligence (mechanical, analytical, intuitive, empathetic), and underlying methodologies (machine learning, deep learning, NLP). These categories help explain the wide range of AI applications today and guide future research and development in the field 1234+3 MORE.
Sources and full results
Most relevant research papers on this topic
4
Artificial intelligence in healthcare: past, present and future
AI is revolutionizing healthcare by enhancing early detection, diagnosis, treatment, and outcome prediction in various diseases, including cancer, neurology, and cardiology.
Rigorous JournalHighly Cited
2017·
3566citations
·F. Jiang et al.PDF
5
Artificial Intelligence in Service
AI job replacement in service tasks progresses from lower to higher intelligences, reducing the importance of analytical skills and increasing the importance of intuitive and empathetic skills for service employees.
Highly Cited
2018·
2734citations
·Ming-Hui Huang et al.6
Brain Intelligence: Go beyond Artificial Intelligence
"Beyond AI" aims to develop an intelligent learning model called "Brain Intelligence" that generates new ideas without experiencing events, improving automatic driving, precision medical care, and industrial robots.
Highly Cited
2017·
1008citations
·Yujie Li et al.PDF
7
Artificial Intelligence in Pharmaceutical and Healthcare Research
AI technologies, particularly deep learning and neural networks, can revolutionize healthcare and pharmaceutical research, improving disease diagnosis, personalized treatment, drug discovery, and epidemic forecasting.
Systematic ReviewHighly Cited
2023·
153citations
·S. Bhattamisra et al.PDF
9
Artificial intelligence for brain diseases: A systematic review
AI has the potential to improve diagnosis, planning, and outcome prediction in brain diseases, but requires comprehensive data and explainable algorithms for practical use.
Systematic ReviewHighly Cited
2020·
114citations
·Alice Segato et al.PDF
10
Artificial intelligence in clinical and genomic diagnostics
AI-based computer vision approaches show promise in clinical diagnostics and clinical genomics, with potential for individualized medicine and risk prediction in complex diseases.
Rigorous JournalHighly Cited
2019·
345citations
·Raquel Dias et al.PDF