It’s 2018, why haven’t you hired a robot yet?

Robots aren’t taking over the world but they might get a job in your business. We explore why that’s a good thing.

Are robots going to take over the world? Probably not, no matter how many times an AI beats a human at chess. It’s easy to predict a Terminator-style future when all we read are human versus robot scenarios.

But in reality, robots aren’t clunky metal figures with glowing eyes and an insatiable desire for human destruction. More often than not they’re a few lines of code or an intuitive interface, helping us in ways we aren’t even aware of.

Of course, there are some standout examples where robotics has crossed into the realm of (almost) being lifelike. Recently, Sophia, the robot took the world by storm. Headlines like “Sophia the robot becomes the first humanoid Saudi citizen” took over the internet, sensationalising the human-looking robot who could interact with people and hold conversations. What most stories failed to report was the fact Sophia was a PR stunt, designed to publicise a tech summit in the Kingdom of Saudi Arabia. It was part of a bigger push to move away from an oil-based economy into technology.

In fact, Sophia is essentially a glorified chatbot, filled with pre-programmed phrases in response to keywords. The distinguishing factor was her lifelike animatronic shell, which acted as a creepily familiar interface.

Having PR stunts like Sophia doesn’t change the effectiveness of robotics though. While it’s naïve to assume that robots will soon be running around like sentient best friends (K9 from Dr Who) or trusted equals (Data from Star Trek: The Next Generation), their various forms are having a tangible influence on how we live our lives.

Research is one of the best examples of this. Codebreaking and translation services were revolutionised by the introduction of AI. By feeding a program 380 languages, scientists in Canada were able to develop algorithms that decoded a mysterious ancient book, which had thwarted people for centuries.

Similarly, scientists are coding AI to improve on what we’ve developed. Physicists in the US are currently attempting to create AI that mimics the way the human brain works – only faster. A machine with the power to think as effectively as a human but with the (almost) limitless power of AI could revolutionise the way we process data, solve problems and make decisions.

It’s important to remember that the faster an AI can learn, the faster it can process data. Currently, we’re producing more data every day and that shows no sign of slowing down. The Internet of Things is creating an extra layer of data that seemingly gets collected and disappears into the ether. Collecting data for the sake of data is a waste of time and money, but processing and analysing the huge volumes of stored information is a mammoth task. So the faster our AI can learn, the faster it can create trends and models we can act on.

This form of AI is called machine learning. Deep learning, a subset of that, can analyse data against thousands of variables, by testing and running it through simulations. Having an AI which learns on the job and could think laterally like a human gives it a chance to achieve more and exceed the limitations of traditional programming.

For example, AI has already been proven to be able to create original artwork, whereas previously only imitation was possible. The jump from replication to creating original content is an incredible shift in ability and one which shows that having a contextual understanding of data can lead to incredible changes.
 

The application of AI in business

We can program AI to do whatever we need, which means its application in a business context is virtually limitless. Two main areas of growth include:

1. Streamline and improve existing processes by capturing and analysing data and automating where appropriate

2. Expand existing product offerings or potentially invent entirely new products

Accenture has predicted that integrating AI into the mainstream could give the global economy an incredible £3.4 trillion boost by 2022. Despite this, many people aren’t clear how they’re meant to make AI fit into their workplace. Automation is arguably one of the quickest wins, but automation for the sake of automating isn’t effective.

Taking the time to analyse how a business runs and what the pain points are will reveal the easiest ways to make effective changes with AI. If your customer service team struggles to prioritise high-risk complaints, then an AI that monitors keywords and prioritises inboxes could improve response times and customer satisfaction. And if your inventory storage and logistics department swing wildly between empty shelves with delayed deliveries to too much stock and not enough storage space, having an AI to monitor and provide alerts could make your system more efficient.
 

When did you last speak to a robot?

Chatbots are becoming a central part of the customer experience. Often businesses are automating their FAQs or troubleshooting sections, meaning customers can ask questions and get relevant responses delivered directly to them.

In East Asia, chatbots are part of business as usual. Chinese chatbot Xiaoice, developed by Microsoft Beijing Research, is one of the most successful and integrates with Weibo, WeChat and even e-commerce platforms like JD.com (comparable to Amazon or Alibaba) and Microsoft smartphones. Being designed as a personal assistant means its function is to help people, a versatile skill that fits many different formats. In a bizarre cultural phenomenon, some people spend hours chatting to Xiaoice, as a coping mechanism for loneliness.

In the Western world, people and businesses exchange more than two billion messages a month on Facebook Messenger. While many have embraced chatbots and automated replies, the countless hours are required to respond to each message. And with constant access to brands through connected devices, it’s never been easier to click and type out a query or complaint.

A team of human customer service agents rarely give a consistent performance. Training, experience and motivation all play into the level of service provided. A chatbot can respond to multiple people at the same time and won’t give an off-brand response because they turned up to work with a hangover, or are getting hangry before lunch. A chatbot can learn from customer complaints and queries, adapting to patterns of speech, behaviour and habits, which can then be fed back to the business and used to improve the programming of the AI behind it.

One of the biggest arguments for automating certain aspects of customer service is that it frees up time from what is usually a labour intensive job. If people aren’t glued to the phone or their inboxes, then they could put their skills to work elsewhere in the business, improving processes or taking high priority cases from their chatbot counterparts.
 

When good AIs go bad

As with any technology, for every altruistic use, there’s a more sinister example just waiting to appear. And there are no surprises that this is happening with cybersecurity. In the same way that machine learning algorithms are boosting cyber defences, they can also be turned to the dark side, helping to break through firewalls and multi-layered security faster than ever before. A vast majority of information security professionals (62%) think that hackers will push weaponised AI into the mainstream in 2018.

Imagine what would happen if your personal assistant AI got hacked and started controlling your linked IoT devices or even started spying on your conversations and digital activity?
 

What’s next for AI

Instead of hurtling toward a Terminator-style apocalypse, think tanks like the IDC are predicting £21 billion will be invested in AI for enterprise-level businesses by 2024, while Forrester says £13.6 million new AI-related jobs will appear by 2028.

But what will set businesses of all sizes apart is the ability to think differently and understand that old systems and ways of working aren’t going to cut it. Create space to collect, contextualise and react to learnings.

What is clear is that there’s soon going to be a race to get ahead in AI. So whether you’re thinking of introducing it to your business or you’re keen to flex your entrepreneurial muscles, there’s never been a better time to get serious about AI.
 
Learn more about how to make AI part of your hiring process and how businesses need to ignore the sceptics and start introducing the Internet of Things.