There is a lot of excitement behind the future of AI. Deep learning is the technology construct that underpins this. It’s part of a broader family of machine learning methods, however deep learning goes a step further than task-specific algorithms and uses neural and belief networks to learn from the patterns and input that it experiences, continuing to learn as it runs.
Though narrow in principle, deep learning as a technology is the closest we have got so far to creating machines that exhibit human behaviour. They are still a long way from our own capabilities.
Deep learning solutions are already being implemented in a number of commercial ways from aiding detection of dialects in translations and image classification to creating better customer service experiences.
One of the most notable uses of deep learning is in Autonomous Vehicles, though this is still at prototypical phase. Rather than use one single model, helping these vehicles get from point A to point B, some deep learning models specialise in street signs while others are trained to recognise pedestrians. As a car navigates down the road it can be informed by millions of individual AI models that allow the car to act.