Artificial Intelligence (AI) has come a long way since its early days as a futuristic science fiction concept. It may still seem like it’s far in the future, yet there are already several types of AI at work today.

With the potential uses for AI being limitless, many can get really caught up trying to figure out the superb one to meet their ever-changing wants and needs. That’s why we’re here to help you understand two types of AI that you should be focusing on and which one is better for you.

Narrow AI

A narrow AI, also known as weak AI, refers to AI that focuses on specific, defined problems. This type of AI is what’s used in many predictive text-to-speech systems.

To illustrate that, if a customer calls a company’s service centre, a narrow AI system might be able to recognize keywords or phrases mentioned that would be useful to the representative (e.g. “My air conditioner is making noises, what could have gone wrong with it?”).

Those detected keywords (e.g. “air conditioner”, “noises”) will then be the factors of consideration for the machine to perform further tasks as programmed, such as to schedule a repairman.

AI Type 2g1
AI replacing customer service.
Source: Annette Miesbach

What does this mean? It means that AI systems can be trained to perform a specific task extremely well, for it does not need to think and process information like humans do. It only needs to know what to do next when a situation is met.

However, be careful! While there are plenty of people out claiming to have AI, it’s usually a simple case of a computer program mimicking a few core human tasks, such as translating between languages, analysing images, and replying to customers.

These kinds of low-level, repetitive tasks can be automated using the Narrow AI with promising speed and accuracy compared to a human being who often makes mistakes and gets bored of a mundane cycle of tasks given.

Broad AI

The opposite is broad or general AI, which allows a machine to learn from its experiences and be creative, constantly learning something new.

AI with human-like intelligence
AI with human-like intelligence.
Source: Ryzhi

For example, if you have a robot that is supposed to be very strong, you can teach it to lift heavy things. That would be the case for broad AI. You can teach the robot to acquire new skills. Broad AI can also use its knowledge to predict what will happen next.

This type of AI can function like a human brain. Besides processing vision and language, it can even go beyond that by performing the thinking and having contextual understanding. With any of the knowledge gained, it will then be applied in another given situation. If it doesn’t work, the machine will brainstorm itself and find a new way to adapt to the new environment.

Comparison: Narrow AI & Broad AI

To put it simply, when a machine does something expected, it’s a narrow AI. When you’re driving your car and the brakes fail, the machine that aids you to stop on the spot and prevent you from crashing? That’s a narrow AI.

In the case of Broad AI, it is a computer that is smart enough to think for itself. By implementing this technology in a car, it will be like those self-driving vehicles that show a wide range of similar human-like common-sense capabilities.

Type AI 1 i 1 1
Self-driving cars with AI.
Source: Jonas Leupe

If you were to think from the perspectives of logic and innovation, broad AI will be the one that covers both of the logic and the innovation that inspires, while the narrow AI is only on the logic part.

In conclusion, it is important to remember that both AI is different. One of the AI is extremely narrow in scope and only capable of performing one specific task. Another AI is much broader and can be trained to perform any task you throw at it. From there, people often mistaken that broad AI is better than narrow AI. Yet, the truth is that it depends on the situation that the type of AI chosen should help you attain your desired outcome.