What is Artificial Intelligence?

AI is a frequently used term – but what is Artificial Intelligence?  Our blog delves deeper into what it is and what all those terms mean…

There are 2 broad types of AI:

  • Narrow AI, or ‘weak AI’, is basically a simulation of human intelligence, which focusses on doing a single task very well.
  • Artificial General Intelligence (AGI), or ‘strong AI’, using its intelligence to solve any problem.

There is more which requires clarifying:

  • Machine learning is where large amounts of data are fed into a machine which is used to learn to carry out a specific task
  • Neural networks are interconnected layers of algorithms which feed data into each other to facilitate the machine learning
  • Deep learning is a subset of machine learning, where the neural networks are expanded with a huge number of layers. This is enabling robots to tackle complex tasks such as speech recognition and computer vision.
  • Knowledge engineering is where a system, or machine, effectively mimics the thought process of a human expert
  • Natural Language Processing is a type of AI that allows computers to break down and process human language
  • Robotic Process Automation is easily programmed software addressing basic tasks
  • Evolutionary computation sees genetic algorithms going through random mutations to evolve the optimal solution to a problem

Machine learning can be further differentiated into types of learning:

  • Supervised learning, where a large number of labelled examples are fed into the system – for example a large number of pictures of a dog.
  • Unsupervised learning is where algorithms look for patterns in data in order to categorise that data.
  • Reinforcement learning is where the system goes through a process of trial and error until it arrives at the best possible outcome.

 

Comment

Liz Reid, MD of RNR Robot Solutions, comments:

“What is clear is that AI is here to stay and will continue to have a major impact on the lives of us all, from driverless vehicles, to facial recognition and surveillance software.  AI has been around for some time – Siri celebrates its 10th birthday this year.  The big questions are about how far it will be allowed to develop, and what controls will be needed to ensure the technology is beneficial”.

Ethical considerations include:

  • Machines becoming so highly developed – outstripping the capabilities of mere humans to the extent that the machines take over
  • The ability of machines to hack into the private lives of humans
  • Do machines have rights? (see our blog which addressed these issues here)
  • Will clever machines take the place of humans in the workforce? (We looked at these issues here)