AI and ML explained... in under 100 words
AI here, ML there – in recent times the terms ‘Artificial Intelligence’ (AI) and ‘Machine Learning’ (ML) have become very popular. They play a prominent role in most conversations about the future of technology, business, the workplace and even humanity itself. It is therefore not surprising that leaders in business and technology are increasingly interested in finding out more about them and their potential applications, benefits and risks.
While the hype about AI and ML may be new, the underlying technologies are not. In fact, their origin can be found over 100 years ago: Linear regression already existed in the 19th century and neural networks were conceptualised in the 20th century. However, advanced algorithms were applied to concrete business problems only in the 21st century, thanks to advances in computational power – processors that are achieving ever higher speeds – and the widespread availability of large amounts of data.
As part of the “Explained in under 100 words” series, we take a closer look at AI and ML and explain the differences as well as their cognitive advantage.
I’m shopping online, I type in a product I want to buy. In addition to this product, other products that might interest me are automatically shown. There is no need for explicit rules; the system figures out the recommendations on its own. This is Artificial Intelligence (AI), an intelligent system that is able to perform human-like tasks such as recommendations.
The system learns how to make recommendations by observing which products buyers are interested in. The more users the system observes, the more it learns about their buying behaviour and the better its recommendations become. This is Machine Learning (ML).
The above explanation is of course simplified and AI and ML have many more cognitive advantages that deserve a more extensive explanation. One key aspect is that the aim of AI and ML is not to replace humans, but to augment their capabilities. As AI is able to tackle routine tasks and increasingly complex non-routine tasks, humans can concentrate their efforts on tasks that have the most added value – those that really need human judgement. For instance, staff deployed in operations do not need to go through every invoice and process it in the appropriate way for the supplier. Instead, they can focus on the more complex ones while the AI algorithm processes the great majority of the invoices – faster, cheaper and more accurately than humans. By learning from large samples of data, AI algorithms can recommend better decisions than humans, discover complex patterns that are hard for the human eye to spot and accurately predict outcomes by taking into account more variables and complex associations than a human ever could.
Most AI algorithms specialise in a very specific task such as recognizing human faces in images or filtering out spam emails. By being trained specifically, the AI algorithm learns to perform the task increasingly well and can outperform humans. However, the field of AI is vast and there are aspects of human intelligence that machines have yet to learn, such as social and emotional intelligence.