Different types of AI
Expert systems
Expert systems is a term used for systems to support decision-making. These systems are stated to store “knowledge from experts”. Expert systems typically follow a rule-based approach and generally lack the ability to adapt and learn from previous data.
Machine learning progress
Machine learning tools are used to identify patterns and relationships in large data sets and are able to ‘learn’ from this data. A common machine learning tool is ‘neural networks’, which predict project success. Off-the-shelf solutions such as TensorFlow by Google and services by IBM Watson Studio have significantly increased the accessibility of these tools. However, research shows that the practical implementation of AI is often delayed due to managers being uncertain how it can be used in their organisation and the difficulty of reusing AI models for different purposes.
Deep learning
As a sub-set of machine learning, deep learning offers a more complex way of analysing data. The ‘black box’ phenomenon is often used to describe the difficulty in interpreting the reasons behind the output of deep learning models.