TechMunch: AI...even the experts are unsure on what the future holds

2017 is already ‘the year of AI’, and we’ve only just finished January.

The world’s elite at Davos this year were actively debating the impact that AI and automation will have on industries, economies and currencies across the world. The technology is being touted as a large piece of the puzzle in achieving the idealistic ‘Fourth Industrial Revolution’ that global leaders are craving. The CEO of Microsoft, Satya Nadella, claimed that AI could be the key to kick-starting economic growth around the world – so no pressure there then!

It appears that AI is not just a sci-fi geek’s dream anymore – it’s really happening, and is now part of the topline agenda for leaders and policy makers around the globe.

There’s a massive appetite for automation in the business community. According to market research company Tractica, the enterprise market for Artificial Intelligence will increase from its current $202.5 million to $11.1 billion by 2024. That’s a staggering increase in less than 10 years.

Data seems to be the key to this progression now. This essentially provides the fuel for algorithms and computer programmes to learn behaviours and adapt to them by assessing a variety of situations. This is something we all learnt ourselves recently during a tech team trip to the Southbank Centre where we heard from TV’s favourite Mancunian science professor Brian Cox and a panel of industry experts discussing the subject of machine learning.

The panel included some of the UK’s best researchers and minds from the field of automation and AI – the likes of Professor Jon Crowcroft from the University of Cambridge, Professor Joanna Bryson, Reader in AI Ethics at the University of Bath, and Professor Sabine Hauert, Lecturer in Robotics at the University of Bristol. So if there’s anyone you want to hear talk about the future of AI, these are the people you want.

The main concern from many – not only from the business community, but also on the panel at the Southbank Centre – seems to be the impact that this advancement could have on human jobs. As an example, almost 3% of working Americans have a job driving of some sort, and AI is tipped to have a substantial impact on the automotive industry. On one side, this could improve manufacturing efficiency and costs for automotive services like Uber, but on the other side serious questions are raised over issues on safety and customer service.

The technology has some great uses in industry – from a live demonstration in the theatre hall we saw firsthand how drones are using machine learning technology to help oil and energy companies maintain and repair remote oil rigs – something that was previously very costly and dangerous for human beings to undertake, but can now be done safely.

Yet the overwhelming motion from the panel was that machines cannot replace humans. Any machine that uses AI still needs to be programmed to ‘learn’ by a human, and we have full control over that. Similar to teaching a person to improve from its mistakes, AI can only gain knowledge by learning from a previous scenario. If anything, there will be an upskilling required to operate and run these machines. The panel also reflected that jobs in the creative industries, such as art, theatre and comedy cannot be replaced since machines cannot reflect the personality or skills that makes us great artists, performers or comedians.

The pitfalls for AI will seemingly lie in regulation, particularly with the defence or arms industries. The panel discussed that, without the right framework in place on a policy level, there’s little to stop someone developing IoT connected weapons, or programming machines to carry out a certain order. This conjures up images of Arnold Schwarzenegger as the Terminator, but the reassuring insight from the panel was that the UK government amongst others are actively looking into regulating the industry quickly.

Regardless of the debate – whether it be on job losses, industry profits, weapon development or economic growth – there is one universal ‘BUT’ which the panel were careful to stipulate across all of this… no-one knows for sure what will happen. It’s all just conjecture at this stage.

So when the next headline pops up in a tabloid newspaper telling us how AI will take over the world , it’s worth remembering this advice from the researchers that are at the forefront of knowledge in this space. The Skynet-style Armageddon may have to wait a little longer.

You can watch the full live show on the Royal Society’s YouTube page.

Ben Rose, Senior Account Manager, Technology