Data revolution

AI and the data revolution

Since the early 1980s, when personal computers were first introduced, we have roughly doubled the amount of data we create every two years. Fast forward 30 years to 2013 and we had created four zettabytes of data which is roughly 4 billion times the size of the average hard drive at the time with 90% of that data created in the previous two years. Its estimated that by 2025 we will be producing as much data in a single week as we did in the first 30 years of the home computer revolution.


That’s not the end of the story. In addition to the volume of data, the complexity of that data is also increasing. Before smartphones the data we created tended to be simple and straightforward, but now with smartphones we are creating videos, tweets, hashtags, likes, social network connections, and the list goes on.

If our traditional methods of gaining insight aren’t struggling with the large volumes of data, they simply aren’t compatible with the likes of video, voice and social media. This problem, however, has been a blessing in disguise – it has given birth to new approaches and solutions such as artificial intelligence.

What is AI?

Through recent years, the buzz around artificial intelligence has only skyrocketed. But too little attention has been given to what exactly it is. Certainly, in the form of robots and machines, it has been the focus of science fiction for decades, being cast in a range of roles from villain to heroes, and even quirky friends. Recently, artificial intelligence has transitioned from science fiction into reality thanks to a revolutionary new technology known as deep neural networks.

Deep neural networks (DNN) forms the foundation of all AI. DNN use layers of artificial neurons to create a miniature brain that specialises in performing a particular task. Although simplistic in function, the networks making it possible are highly complex. Compare this to the human brain, commanding a plethora of simultaneous functions and you would have hundreds and thousands of networks that even experts today don’t fully understand. This layered architecture means computer systems can process information and make decisions not unlike us humans.

Currently, DNNs only excel at simple and very repetitive tasks. For example, DNNs are very good at automatically identifying and distinguishing between different objects in images. DNNs have revolutionised the fields of computer vision, natural language processing, and data analysis and is continuing to branch out into more diverse fields.

Artificial intelligence is already revolutionising many industries across the world, from healthcare to finance. With 47% of organisations reporting to be investing in automating tasks and 30% of respondents explicitly using AI and robots to perform routine tasks or augment human skills, this trend will only grow in the future.

AI invesmtnent

No Comments

Post A Comment