How close is artificial intelligence (AI) to changing how we see the world?

The technology made a grand entrance into the world of mixed martial arts recently, in a way that has the potential to change how the story of a fight is told.

At re:Invent, an Amazon Web Services (AWS) conference that was held in Las Vegas last month, a mock sparring session between two fighters used analytics and AI.

Gloves, cameras and mats equipped with sensors measured everything from punch strength to the stress levels of fighters and even their family members.

The vast amount of data that can be extracted from such sensors can be interpreted for audiences in real time, giving them entirely new information and interpretations about a fight.

Mati Kochavi, one of the founders behind the tech, had this to say: "Shouldn’t sport be told in real time, with real data, with real information, with real insights and real emotions?"

Also at re:Invent, online travel company Expedia took to the stage to share how machine learning has improved the customer experience of its 600m users.


Expedia CEO Mark Okerstrom says the company is able to handle large amounts of data on the fly, servicing 1.6m customers daily. Expedia handles 750m searches a day, and is able to scan 3m hotels in a second for personalised results.

When one considers applications of technology that range from voice-powered personal assistants like Amazon’s own Alexa to autonomously powered self-driving vehicles with predictive capabilities it becomes easier to see how AI could allow the machines to take over. Computer-aided interpretation of medical images, heart-sound analysis and even the composition of classical music by means of AI are just some of the advances in the field.

AWS recently unveiled at least five new machine-learning tools, many of them intended for businesses. They let companies develop their own algorithms and allow software to translate in bulk, compile libraries of data, analyse video, extract key phrases from big data and transcribe from audio files.

Matt Wood, GM of AI at AWS, tells the Financial Mail that AI and machine learning, specifically robotics, lead to greater business value, and allow businesses to grow at a faster rate.

"If you evaluate Amazon’s growth by any metric, it’s been significant; and you can attribute some of that to investments we have made in AI from robotics."

Amazon has been investing in machine learning for over 20 years, and has thousands of engineers that are experts at it, says Wood.

On the hardware front, the company has also created the world’s first "deep learning-enabled" video camera. Called DeepLens, it uses neural networks to learn and make predictions.

Wood says the camera can be programmed to do "almost anything you can imagine". It can be taught, for instance, to recognise a car licence plate. This means that when you pull into your driveway, it can open the garage door automatically. Or you could programme it to send you an alert when your dog gets onto the couch, says Wood.

Closer to home

Wood says that the top three industries that could benefit from AI on the African continent sooner rather than later are transport (autonomous cars), health care, and manufacturing.

"We have a lot of customers working on medical imaging, for example for the early detection of tumours," says Wood.

And the manufacturing industry could improve their processes through automation, he says, citing an AWS customer in France who is doing that for its silicon computer chip business.

"Using a machine learning library that we introduced with Microsoft a couple of months ago called Gluon, the system automatically scans the surface of the silicon, looking for defects, enabling [the company] to improve the quality and output of the manufacturing process dramatically and reduce costs."

The power of data

In the past, companies had a number of constraints on the amount of data they could collect, how much computation they had access to or what hampered databases and analytics.

All that has changed with unlimited storage.

Once those constraints start to fade away, Wood says, the true value of the data comes into play.

"You are able to move from not having any machine learning to having very sophisticated machine learning in literally a week."

• The writer was a guest of Amazon