There has been significant fanfare as the evolution and rapid progress of AI technology shows the potential for industries to move from established automation practices to the digital frontier of artificial intelligence through application of deep learning. Self-driving cars, competitor insights and translation are all examples cited in this interesting article from Wired.
In practice, the heavy demands of computational processing required to fuel this degree of analysis are reaching a barrier on cost effectiveness - with other implications such as the carbon footprint of such practices also looming large. Researchers at MIT are currently preparing further case studies on this topic.
This is a great example of where the shiny 'brochure-ware', as I call it, surrounding digital innovation sometimes falls short when it comes to the case for investment and delivering tangible benefits to businesses. The Covid-19 crisis has led to a necessary increase in scrutiny from cash-strapped companies as they choose where to invest their funds in technology and deliver new capabilities.
The party is far from over for Artificial Intelligence and Deep Learning as progress continues to excite and delight many of us. However, perhaps the pace of change may take a slight correction while the supporting technology catches up, in order for broader adoption to take place.