The AI revolution is often spoken about in hushed, fearful tones. Its reputation hasn’t been helped by Hollywood storylines depicting artificially intelligent robots destroying our future and the world as we know it. For the technology to hit the mainstream, tech leaders must work to allay the public’s fear.
It’s already here
It will come as no surprise to many people that AI is already a large part of everyday life. Much of what we do today is augmented by artificial intelligence; our smartphones are made smart through AI – those gimmicky animojis that are doing the rounds on Apple phones, for example. Smart home assistants like Alexa and Google Assistant work because of machine learning. Even Tinder is powered by algorithms, and Spotify’s recommended songs the same.
In the business world, organisations are increasingly turning to AI to make processes more efficient and cost-effective. PayPal uses AI to detect fraud, law firms use it to build cases and predict success, and it’s allowing marketing approaches to be much more targeted. Although many products aimed at businesses appear to use AI as a marketing buzzword, eventually it’ll become so commonplace it won’t be a USP at all. We’ll come to expect it to power every piece of software we encounter.
To get there, however, there are hurdles to overcome.
Asking people to trust AI is a tall order when they hold fears of losing their jobs to the technology. To combat this, expectations must be reset – far from taking our jobs, AI is going to improve them, as well as creating new roles. It’s true that manual labour and low-paying jobs are likely to be automated, with those workers needing to reskill and switch careers. However, for the vast majority of the workforce, their day-to-day roles will become more creative, engaging and strategic. Machine learning will pick up a lot of the legwork that nobody really enjoys – filling out spreadsheets, for example. Learning this, people may be more inclined to embrace and trust AI.
Of course, showing them the benefits of AI in their role will cement that opinion. There are many AI tools on the market today, so pick one that aligns with a departmental or business goal and then find an AI tool that’s relatively quick and cheap to implement. Once it’s working, show the rest of your organisation the results and they’ll rapidly get on-board with more ambitious plans.
Artificial intelligence with truly ‘human’ characteristics will be a long time coming. There are limitations with current technology – even its more advanced subset, deep learning, doesn’t fully reflect human intelligence just yet. Skills like empathy, creativity and critical thinking may never be mastered by a bot.
Luckily for us humans, this means that we’ll always be required in some capacity. Human intuition will be called upon to analyse an AI’s insights, for example, or to help humans and robots work collaboratively. Deloitte has proposed that factories will hire Robot Teaming Coordinators to complete that exact task.
As AI becomes more advanced, the tasks it is able to complete will too. This raises some ethical questions, as researchers tackle the thorny issue of programming ‘right and wrong’ into AI. Given that our sense of right and wrong is subject to cultural, geographical and experiential influences, this will be extremely complex. The consequences of getting it wrong can be dire.
In the future, machines will have to make snap decisions that’ll dramatically impact human lives. A self-driving vehicle, for example, could find itself in a real-life trolley problem. A care robot may have to decide whether to force a non-compliant patient to take life-saving medication.
Another danger surrounds current algorithm development and the risk of introducing (and worsening) unconscious bias. Google fell foul of this when it developed an image recognition algorithm that tagged black people as gorillas. The problem with unconscious bias is that we often don’t know what presumptions we have. If these are introduced to an AI, then our existing biases and inequality will be made worse. A world run completely by algorithms may see more women sidelined for jobs, for instance.
A potential solution is to have diverse teams working on AI, who can challenge their colleagues’ biases and theoretically prevent these from entering an algorithm. Unfortunately, female, BAME and other minority AI researchers are hard to come by in the current environment – a very prominent issue in itself. Another option is to have an AI justify every decision that it makes. Avoiding black box AI – where you don’t fully understand how it has produced a certain output – is key. Human oversight, therefore, must be ever-present.
Questions left to answer
We don’t have to imagine a future with humans and AI working alongside each other – it’s already a reality. For AI to advance however, humans are left to address some complicated questions around the interactions that they want to have. The technology will have an increasing impact on all of our lives, but for the smaller nuances, like working together and empathising with colleagues, the ball is still very much in our court.
AI can present huge opportunities for you and your organisation, no matter what industry it is you work in. If you want to find out more about how technology can benefit you, get in touch with us by emailing us at firstname.lastname@example.org or calling 01225 220155.