Over the past few years, we have seen machines become increasingly intelligent. Real-time voice translation programmes, smart thermostats, and inbuilt sensors are making the objects around us more responsive and intuitive.

In fact, at this year’s CES convention, the technology world pledged its biggest ever commitment to the Internet of Things (IoT). Samsung, for example, promised that in five years’ time, every piece of hardware it makes will be IoT-ready. Google’s smart thermostat, Nest, announced 15 new hardware partners, whilst many of the devices on display were already certified to meet Apple’s HomeKit IOT standard.

However, such moves towards a future founded upon intelligent machines haven’t been without its critics. Stephen Hawking has predicted that at its full development Artificial Intelligence (AI) will be “the biggest event in human history” but “might also be the last”. Elon Musk of Tesla Motors and SpaceX has urged more investigation into the area, joining with leading futurists at Google and the Future of Life Institute in calling for further research into ways to avoid the potential pitfalls of AI.

Beyond the scaremongering

Machine learning is one way that we are approaching AI, and it is not particularly helpful to see the apogee of this technology in terms of just recreating human intelligence. Computational neural networks, which have been in development since the 1940s and recently experienced a resurgence in popularity, only loosely mimic the way the brain actually works (which in itself largely remains a mystery to mankind).

The fact is that, with an increasing number of sensors being put into IoT devices, we need machine learning. The amount of data we are producing is increasing exponentially, and we as humans simply do not have the capacity to understand it all. Rather than being in competition with man, we actually need increased computer intelligence in order to aid us in processing so much information.

Machine learning and the future of search

This application of machine learning will fundamentally change the way we access information. When search engines unlocked the potential of the World Wide Web, what was once a vast collection of inaccessible text pages suddenly became an infinite resource at peoples’ fingertips. Search engines indexed these web pages, and specialised in returning text-based search results. However, as we move into the age of IoT and wearables, this text-centric way of organising information becomes less appropriate.

Wearables, specifically, are constantly exposed to our human interactions, and machine learning software will allow them to actually understand the things they are seeing and hearing. From this, they will then present us with further information or search results in far more human-friendly forms: with speech, conversations, images and video.

Despite great advances, the history of AI has been plagued by naysayers. But while it is important to sense-check major developments in line with Hawking and Musk’s warnings, we must stop thinking in simplistic terms of a dichotomy between humans and machines. With scientists telling us that we’ll need to invent new units of measurement to quantify the data we produce in 2015, machine learning is the only viable reality – and one that has been ushered in by our very human demands.

Charlotte is co-founder of Sense, an intelligent recognition platform for wearables and other smart devices. The software forms part of a wider move to step away from text-centric ways of interacting with information. 

Show Comments