The development of artificial intelligence over the past few years has seen many businesses explore ways to apply it to their business in a way that will benefit them and augment their existing systems and structures.

However, it has not necessarily been a straightforward path to reach an era where businesses look for AI training courses to get a system which suits their needs.

Possibly one of the best examples of this point is the fact that one of the breakthrough moments of AI was when one was trained to recognise pastries before it was later used in a cancer-detection system.

In the mid-2000s, Japan went through a patisserie revolution, with huge lines forming at many small artisan bakeries in the country’s major cities.

The problem was that Japanese customers did not like wrapped pastries because they looked less fresh, but as a consequence, this means that there was no place to stick a barcode and means that checkout staff needed to tally up as many as 55 different types of bread products by hand.

The solution came from an early AI company called Brain, which developed a custom algorithm for training an AI system to recognise and learn types of bread with relatively limited amounts of useful data.

The system had an accuracy rate of 98 per cent across the 55 varieties, aided by the development of a backlight which stops shadows or bad lighting from affecting the results.

The success of BakeryScan AI was significant and led to a cultural watershed when it came to AI as a concept. It was a successful, practical use of the technology, one that became even more popular when it was adapted for use in cancer detection.

It is not the same type of system as a modern deep-learning system, which usually comes to the right answer for reasons that are nigh impossible to fathom, but it was a highly effective proof of concept.