Deep learning presents the organisation with an entirely new realm of unexplored wonders, and impressive potential, by Johan de Villiers, Managing Director, First Technology.
Deep learning defines a machine’s ability to simulate the vast potential of the human mind. It utilises algorithms, software, hardware and computational power to learn from the experiences and data which are fed into its database, and takes these learnings to uncover solutions, create simulations and so much more. It is an idea which has excited brilliant minds since as far back as the 1970s, but has never quite found a foothold thanks to limited computing power and data capacity. Today, however, those limitations are not quite as extensive as they were 40 years ago, and deep learning has become an accessible tool which the business can harness to achieve any number of impressive ends.
Today, deep learning is far more than an idea. It is a reality which has already permeated business and marketplace. Google, the search giant which is on every screen, uses deep learning in its voice and image recognition algorithms. The company has also recently purchased DeepMind, an artificial intelligence business which has created algorithms that allow machines to learn independently. DeepMind’s fundamental goal is to use deep learning to solve the biggest challenges faced by humanity. Just think of the benefits to the business should it be able to use deep learning to tackle issues which impact its sustainability, profitability and growth?
Predicting your outcomes
This is the potential of a powerful artificial intelligence capable of analysing business data at such speed and with such accuracy that it can show where the business is failing, identify issues or find more efficient ways of handling projects or performances.
Netflix and Amazon are also examples of deep learning success in action. Both have used recognition algorithms to assess customer preferences to improve experiences and engagement. While the results of the current algorithms may not always be perfect, they are close enough to make the technology worthy of the title. In the book, The Netflix Recommender System: Algorithms, Business Value and Innovation, Netflix reveals the algorithms which make up their system and how they influence the business. What is really relevant here is how deep learning has been used to improve Netflix audience engagement and retention as well as giving the company room to experiment with algorithms and data to define new business horizons and methods of engagement.
However, in spite of all these achievements, deep learning is not quite at the point where it will match the capabilities of the human mind. It’s fast and its interesting, but most applications of deep learning are not sophisticated enough to mirror the brain accurately. Think of a person walking to work – they cross the road and instinctively check for cars, obstacles, dangers, assessing the overall nature of the situation. A machine isn’t quite there yet and could very well walk into an oncoming car.
Interesting, it looks like this limitation is soon set to change. The Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory has created an algorithm which can predict the future based on a few images in a scene. This is, in every sense, a door opening to a whole new world. Imagine the possibilities for the business if a machine could instantly assess a situation and determine the best outcomes?
Deep learning is so much more than just artificial intelligence or machine learning, it is the ultimate expression of the two technologies and their capabilities. And it is something which is set to play a major role in influencing how businesses are run and lives are transformed.
Until next time, thank you for your continued support of First Technology.
Johan de Villiers Managing Director