How AI and Automation will disrupt your industry in 2018
Let’s be clear, the conversation about Artificial Intelligence (AI) today is worlds apart from the dystopian images we see depicted in movies such as Blade Runner, I, Robot and 2001: A Space Odyssey. Real developments have been far less glamorous. Nor is the true representation predicated on a physical walking talking robot. This day may come, but today, the truest application of AI in industry revolves around process automation. It is an already dynamic and which promises rapid growth in the immediate future due in the main to the maturity of three key aspects.
How Deep Do? You Want to Go
Two of the most eye-catching moments in the development of AI are the victories of Deep Blue in 1997 and AlphaGo in 2016. Whereas Deep Blue’s victory was a triumph of PR, it was due more to the power of brute strength than intelligence; AlphaGo’s performance displayed much a deeper cognition, derived by teaching itself the game using real-life examples and games against itself.
AlphaGo had no specific programming regarding how to ‘play’ ‘Go’ but merely utilised a combination of neural networking and ‘deep learning’ / reinforced learning to teach itself how to perform better the task. Most ‘Narrow AI’ being developed currently tends to adopt this strategy. Deepmind, the company behind AlphaGo was bought for $500 million in 2014 by Google and has shown brilliance with several significant examples of AI.
AI is Already in Effect
Although the idea is nothing new, the crystallisation as a sophisticated, practical application is relatively recent. Along with deep neural networks developed (which mimic real human neural pathway development in the brain), computing capacity in both power and memory, and the tremendous amounts of data generated and captured have allowed a real leap in state of the art.
Naturally, the ‘Network effect’ players (Google, Baidu, Amazon, Apple, Facebook, Microsoft) dominate the technology due to their access to the unencumbered sheer volume of data and the enormous resources at their disposal. Evidence of this can be seen in your home and on your smartphone every day. Siri, Alexa, Cortana, Google Now as well as Netflix targeted recommendation engine and Amazon’s suggested purchases are examples of the power leveraged by these global behemoths. The apps are pervasive but successful.
Why Does This Affect My Industry?
As we enter the 4th industrial revolution, we must be aware of the impact of AI on the global workforce. AI will engender a dramatic change in the way we work. It is more important though to focus on activities which may be automated rather than whole occupations.
McKinsey reports that only 5% of jobs can be fully automated, but for activities performed within a job, 45% can be automated with current technology, rising to 58% once AI can understand ‘natural language’ to an average human level.
In fact, 60% of jobs could have at least 30% of the role automated to some degree. This proves to be more of a blessing than a curse, as it frees up capacity for workers to do ‘more important things’ driving productivity increases. 65% of executives say AI will help humans and machines work together to be stronger using both artificial and human intelligence. 
For a technology which could have a purported $15.7 Trillion contribution to the global economy by 2030, this is not an aspect of your business you want to neglect.
Predictions for Industry Disruption
We have looked at the world-view and the precious few but how will AI affect you and your industry. It must be noted that there is no magic bullet which will provide the answers, AI requires large scale integration, setup time and costs, but the evidence suggests that early adopters of AI enjoy significantly higher profit margins against the industry average across all sectors analysed in the McKinsey Global Institute survey.
The industries with the largest amount of potential for automation according to McKinsey are resource extraction; construction; accommodation and food services and technology media and Telecoms. However, PwC claim industries where AI can but put to practical work can be ranked with Healthcare, Automotive, Financial Services, Transportation and logistics, technology media and telecoms and retail and consumer as areas where the most significant impact can be made.
AI in practical application
Indeed, these predictions are useful, but real-life examples are being deployed which can have a much more tangible understanding for you and me. The UK does actually show a leading edge in this regard, definitely pioneered by the work of Deepmind, but adopted readily by industry leaders such as Ocado, Tesco and even the NHS.
As an example, Deepmind, in conjunction with the University of Oxford managed to develop a lip-reading system which comprehensively outperformed a professional by annotating 46.8% to 12.4% of words without any error.
Ocado uses AI in but its logistics and customer interfaces, utilising actual routing robots to optimise navigation through its warehouses and factories; applying demand forecasting to its stock and product levels and initiating suggested purchases for users based on their navigation and previous purchase histories.
One of the most prevalent uses of AI currently and only continuing to grow is in the use of ‘Chatbots’. These interactive widgets are being incorporated into Facebook via Messenger as companies aim to reduce the strain generated by phone/email interactions. One example of this is the NHS 111 helpline which is building a chatbot which is capable of giving an automated diagnosis following the disclosure of patient symptoms. This has the potential to alleviate tremendous pressure from an already time pressured service.
Virgin Holidays worked to combine the use of Behavioural insights and AI by outsourcing its control of all subject line content in its marketing emails. By utilising reinforced learning AI they found they could improve the content of their subject lines, resulting in a 2% jump in open rates.
Similarly, to Ocado, Tesco’s have employed AI across their business including driver routing, customised price drop alerts and interactive recipes which match/link to products users like or purchase.
These are some tangible examples, but there are many more generic situations where AI can and will be employed intelligently to disrupt the way traditional business has been done. The Banking industry is focussing on process automation for simple transactions and fraud detection according to users known purchasing habits.
Dynamic pricing, production recommendations and smart email categorisation are all simple but effective ways to drive improvements through increment. Wherever there is a defined business process, there is the opportunity for automation to be applied, and AI can free your organisation to provide greater value to you and your customers.