In the last five years, AI has landed. Developments in mobile, hardware and software innovation have ensured that artificial intelligence is more than just an industry buzzword. Businesses have started to benefit from its ability to solve general purpose problems and mathematic calculations, while headline-grabbing achievements have captured the attention of the wider public. Just last year, AI defeated the Korean grandmaster Lee Sedol at Go – a very complex 2,500 year-old-game requiring intuition as well as skill. Similarly, AI is now being developed with ‘imagination’ and the human ability to construct plans.
At this stage the AI industry is nascent. We’ve seen a number of successful and award-winning solutions, but we’re really just getting started. The challenge now is to ensure further development in the software and hardware that supports AI, domain expertise, data processing capabilities, and IoT architecture, to truly drive the AI industry even further forward and ensure that the right applications of the technology are developed for each industry.
However, there has been some debate around whether AI will become too developed, too quickly. Only recently, technology leaders highlighted concerns around AI in weaponry, whilst Elon Musk went as far as to say that “AI is the biggest risk we face as a civilisation”. At this point, I’m excited to see how this technology is being applied and the results which it can deliver. But with further advances and innovations, we will also need to see better handling of the technology.
Don’t stifle AI before it gets started
AI is not just about statistics and numbers (early machine learning solutions), it brings automated reasoning and digital learning to the masses. In time, it will be dominant across all industries and have profound implications on what we do today, bringing new market leaders with it. In turn, this will create a need for new skills and education (around cognitive capabilities, infrastructure, new roles, platforms, types of applications and an understanding of human-machine interaction). Crucially we need to ensure that appropriate safeguards and regulations are put in place, with a clear understanding of how AI and its decision-making process works. This is particularly important when we start to talk about scoring for safety or assigning rights or costs to an individual. Consumers and businesses alike will require assurances before AI can and should become mainstream.
What’s essential is that we don’t stifle AI before it really gets started. Though I agree that caution is required, I personally hope that AI will be an open, inclusive industry. For instance, we need to ensure that there are variations in how AI is being applied across industries – the approach to smart cities or healthcare will be very different, for example – to realise all the potential benefits to companies of different sizes. To really enable innovation and unlock more capabilities, competition is needed across the industry.
I believe that leaders in AI will come from new markets, those driving digital transformation rather than solely from those looking to solve specific problems. From start-ups located in ambitious hubs such as Japan or Israel, to larger companies in established areas such as Silicon Valley in the US, around the world we’re seeing these exciting developments which will continue to drive innovations in AI. Having this breadth of approach in AI will be crucial in driving the industry forward.
This was part of the reason why we explored, with the help of experts at the Institute for the Future, what business will look like 10-15 years into the future. As mentioned, there tends to be two extreme perspectives about the future: the anxiety-driven issue of unemployment or the optimistic view that technology will solve all of the social-environmental problems of the world. The reality is the future will likely fall somewhere in the middle. As it has already done so in the past 10-15 years, technology will again reshape the world by 2030 and these changes will present new opportunities for humans.
Expectations of AI are high. Yet, solving environmental issues (i.e. smart cities) or finding the cure for disease will take more than one expert algorithm. Industries must digitally transform to take advantage of its promise. In doing so, they need to recognise the concerns which exist but work within safe confines of data management and control, while developing technology which will change the world.