Artificial Intelligence
What is AI?
Artificial Intelligence infuses machines with intelligence. In other words, AI is a science, trying to create machines (software, robots, computers) operate intelligently similar to humans. The operation behind AI uses a system of algorithms, large amounts of data and processing power, multiple disciplines, and AI sub-fields.
The goal of artificial intelligence is to create intelligent systems that learn and behave like humans, but also instill cognitive comprehension and rationality.
The application of AI have many benefits and positively affects multiple disciplines. The following are some of the advantages of artificial intelligence:
- Identifying patterns and trends
- Fast and efficient decision-making
- Unbiased reasoning
- Available at all times
- Reducing human errors
- Performing repetitive tasks
- Reducing human risk by performing harmful tasks
- Promoting innovation
Types of AI
AI can be categorised into three different types: Narrow AI (also Weak AI), General AI (also Broad AI or Strong AI), and Superintelligence AI (also Super AI) – with abbreviation ANI, AGI, and ASI accordingly.
Narrow AI refers to a singular function, aiming at a specific and defined task, hence the name Narrow or Weak. This type of AI is the most common one, applied in almost every industry nowadays. Some well known examples are speech detection, face recognition, recommendation systems, disease detection, and many more.
General AI refers to a solution that can be exercised in various contexts, with the aim of creating systems that can behave and think similar to humans. AGI is what scientists have being trying to achieve for some time, but we are not there yet. The closest examples of this type of AI currently are self-driving cars, some chatbots, and expert systems.
Super AI refers to machine intelligence that can surpass humans by far, aiming to exceed human cognitive performance in all areas of expertise. This type of AI is only hypothetical, and does not seem plausible that it can be achieved in the near future. Some AI specialists predict that humanity will accomplish Super AI by 2040-50.
Functional Types of AI
The above split of AI types are categorised by capabilities. The following are the four functionality types of artificial intelligence.
Reactive Machines do not have memory, and as such, can do not learn as well as only perform basic functions, providing output from the input they have. Completing particular tasks systems are examples of reactive machines (e.g. IBM Blue chess playing supercomputer).
Limited Memory refers to systems continuously storing limited (but enough) data in order to make decisions and provide predictions. Self-driving cars for instance, use sensors to identify civilians and other objects.
Theory of Mind is the integration of AI systems with emotions and thoughts. These AI systems do not completely exist yet, but the closest to this type of AI can be observed in robotics (e.g. Sophia humanoid robot) and autonomous cars.
Self Awareness, as the name suggests, refers to AI systems that have self awareness and independent intelligence that surpass the human mind. This type of AI currently only exists in films and stories – i.e. only hypothetical.
Applications of AI
The development of AI and realisation of its benefits have created a wide usage of artificial intelligence around the world. Nowadays, AI operates in most industries and the following are currently some of the most common cases.
Education
- Development of smart content
- Personalised learning for students
- Learning assistance
Robotics
- Inventory management
- Goods robots in warehouses
- Mechanical tasks
E-commerce
- Fraud Detection
- Recommendation systems
- Chatbots
Healthcare
- Diseases detection systems
- Virtual Health Assistants
- Medical Records Management
Marketing
- Content management and branding
- Products and services personalisation
- Personalised advertising
Data security
- Fraud detection
- Threat detection and prevention
- Data flaw identification
The list of applications of artificial intelligence engines is almost endless. There are many more utilisations of AI – e.g. agriculture, lifestyle, automotive, transport, social media, gaming, finance, astronomy, and so on.
Next: Perceptron
