What is artificial intelligence?
Artificial intelligence (AI) machines simulate human intelligence and can think and mimic their actions. It has human traits of learning, adapting, and problem-solving. AI can rationalise from a range of given options and shortlist the most effective way to solve a problem.
AI technology is growing at a fast pace, and many of our daily activities and decisions are already influenced by it. A leading research firm predicts that AI could contribute $15.7 trillion to the global economy by 2035. AI technology works over a layer of other technologies, such as machine learning, big data, analytics, and annotation.
How does AI work?
AI works on several underlying technologies that process a vast amount of data. Some of the technologies are –
- Big data – Big data is a vast amount of information from distributed systems within a company. It also includes user-generated data, online data, product/service reviews, and much more. The purpose of accessing this big data is to consolidate it and turn it into meaningful information that the system can analyse and feed into ML systems.
- Machine learning – Machine learning (ML) is the underlying technology that gives AI the ability to learn. It uses algorithms to discover patterns in the data that humans feed into the system.
- Annotation – Annotation is the process of labelling images, such as that of vehicles, humans, etc., so that the system can recognise them. The four major types are data annotation or natural language processing (NLP), video annotation, image annotation, and audio annotation.
Artificial intelligence applications
From financial transactions to self-driving cars, AI is already becoming a part of our lives. Some common applications are –
- 1. Banking fraud detection – AI can detect fraudulent transactions before they do significant damage. It can detect patterns and scan a vast number of transactions that are not possible for a human.
- 2. Cybersecurity – AI can easily detect anomalies in the system and predict possible threats from hackers.
- 3. Customer support – AI is automating most online customer support through voice responses that learn from customer interactions.
Types of artificial intelligence
The types of AI demonstrate their capabilities in performing highly specific tasks and thinking like humans. The main types of AI are –
- 1. Reactive machines – These AI machines process a vast amount of information to produce a pre-defined output. For example, processing a customer’s complete Netflix history and suggesting movies that match his/her preference.
- 2. Limited memory – This AI uses deep learning to train itself and improvise the way a human brain does. Limited memory AI learns from events over time to become better at predicting outcomes. A good example of this is self-driving cars that learn about other vehicles, traffic lights, bumps, and curves on the road.
- 1. Theory of mind – While this AI is currently a thing of the future, it could perceive emotions, thoughts, and behaviour. It could understand intentions and predict behaviour as humans do.
- 2. Self-awareness – This AI would have a sense of self and conscience like humans. It will understand emotions and can predict others’ feelings.
However, we are still a long way from the last two types of AI.