Business is on the brink of an automation revolution as artificial intelligence or AI transforms the way we work. AI powers virtual assistants from Apple’s Siri to Amazon’s Alexa and is the essential ingredient allowing Netflix to recommend films. AI systems such as IBM Watson are transforming the way banks and insurers assess risks, rewards and investments, and are helping doctors to analyse data about patients.
But the birth of AI is proving challenging for many organisations. It raises some complex questions about how the use of AI technology will change the way companies do business. Which areas will be affected? What are the first steps to take when integrating this technology and who should take them?
At at the heart of transformation
Some companies are starting to experiment with AI while others are unsure of the next steps to take. Josh Sutton, global head of the AI practice at consultancy Publicis.Sapient, gives the example of two global banks. At one, a senior executive told him the organisation is laying the foundations for deploying cognitive computing, an AI technology which uses machine learning to teach computers to mimic human ways of thinking and undertake human tasks. The bank is ready to start experimenting with this to find ways of driving down costs and serving clients.
AI needs to become something on the CEO’s radar – it is not activity that lives within one of the traditional silos of an organisation
But at a second global bank, the board of directors views AI with anxiety. They recognise the organisation is way behind in AI, but struggle to decide which executives in the business should lead the implementation of the technology.
Mr Sutton says this paralysis occurs because AI does not fit neatly into any specific area of responsibility; it is not exclusively for the chief information officer, the chief marketing officer or even the chief operating officer. Rather, it needs to be part of an overall business transformation driven from the top down.
“AI needs to become something on the CEO’s radar – it is not activity that lives within one of the traditional silos of an organisation, any more than being digital is something that lives with the chief information officer,” says Mr Sutton. AI needs to be approached at a strategic level, he adds, as something that will transform the whole business.
Forward-thinking companies are handing responsibility for implementing AI to senior members of staff such as the chief innovation officer, who will evangelise for the technology across the business. More conservative businesses are bringing in management consultants to create a strategic plan and help implement it.
There is a sense of urgency to this. The lesson of recent years is that businesses need to disrupt themselves before a technology startup or innovation comes along and disrupts their industry from the outside, says Paul Chong, director of Watson Group for Europe, the Middle East and Africa at IBM.
Applying the technology
Mr Chong says some companies are looking at implementing IBM Watson’s AI capabilities through a “competency centre” approach, where they create a structure for working out how AI will affect different areas of the business.
IBM Watson is working with one global company which has put AI implementation under the remit of the chief technology officer, who is managing the competency centre. This starts by laying out a vision of how AI will broadly transform the organisation. Then a series of seminars and events run across the business to work with employees on how it will impact different departments. The idea is to create a hub of expertise in AI across many different areas of the business, though this knowledge doesn’t need to be highly technical.
“You don’t necessarily need machine-learning experts, but you need to know how these types of technologies can be applied. We’ve trained people to understand what makes a really good application of Watson or AI and what is not so good, what is applicable for today and for the future when the technology has advanced,” says Mr Chong.
Meanwhile, Frank Palermo, executive vice president of global solutions at technology consultancy Virtusa, says executives need to know which jobs can be delegated to AI and which ones will always need the human touch. The technology is developing from “weak AI” focused on narrow, mundane tasks to “strong-AI” applications that use sentience and applied intelligence. This will transform the way enterprises manage their software. Strong AI will go beyond merely supporting employees working on customer relationship management and other software tasks; AI will start doing a part of those jobs.
Mr Palermo believes that every task now undertaken by humans will have the opportunity for AI input. “I’m a big fan of assistive technology, human plus machine is really where the benefits come from rather than machines replacing man or man fighting against machines. It is the combination of the two that will become really powerful,” he says. The future for AI will lie in how the technology interacts and co-operates with humans, rather than replacing them, he adds.
Boards of directors will need to work hard to explain the benefits of AI to employees and customers. Otherwise they could face a backlash against the technology as the hysteria about AI’s threat to jobs and customer service reaches fever pitch.
APPOINTING A ROBOT TO THE BOARD
When Deep Knowledge Ventures (DKV), a venture capitalist firm based in Hong Kong, announced it was appointing a robot to its board of directors, it made headlines around the world.
This may have been little more than a public relations stunt, but it also raised some serious questions about the role of artificial intelligence (AI) in boardroom decision-making.
DKV, which invests in companies focused on age-related diseases and regenerative medicine, said the algorithm, called Vital, would make investment decisions by analysing large amounts of data. It would get to vote on whether to make an investment, just like other members of the board.
Observers quickly dismissed the move as a bid for publicity, claiming it would be incompatible with Hong Kong regulations as board members need to be accountable for their decisions and bear liability for any losses.
But the announcement played into the idea that computers could one day assume human responsibilities and take part in corporate decision-making.
The appointment of Vital raises the question of how much power should be delegated to AI. Algorithms are already tasked to make investment decisions on stock markets. However, while Vital and every other algorithm may make recommendations, ultimately the decision on whether to follow them is a question of human agency.