Unlike some technologies that represent a "single point" of innovation – a better mouse or printer, for example – platform technologies create an entirely new foundation, one that enables countless innovations and creates enormous value. The platform technologies to watch for are those that exponentially reduce the cost of something that is useful to a huge market of buyers and sellers: the steam engine (power); railways (moving heavy industrial goods and commodities); electricity grids (powering and using machines); the telephone (synchronous verbal communication); and the internet (data communication) serve as just a few historic examples.
Artificial intelligence – and specifically machine learning – is poised to be one of history's greatest platform technologies. What the steam engine did for physical tasks, AI will do for cognitive tasks. AI is a way of creating intelligent software and machines wherein algorithms do not need to be explicitly programmed. Essentially, you feed an algorithm data and tell it what you want, and it develops the model itself – for example, an autonomous-driving algorithm that learns to drive by studying millions of hours of human driving data, or a natural-language-processing tool (or chat bot) that predicts how to best respond to a customer by learning from thousands of previous conversations. As Andrew Ng, former chief scientist at Baidu, has said: "Just as electricity transformed almost everything 100 years ago, today I actually have a hard time thinking of an industry that I don't think AI will transform in the next several years."
There is a growing consensus that AI will have a significant impact on industries, governments and societies around the world, which raises some questions about how best to manage the coming change. The clear lesson from history, however, is that it is better to embrace the next platform technology than be a victim of its disruption. But how? What action plan should your organization build to prepare and capitalize on AI?
1. Rally the executive team, the board and the organization by articulating your vision and educating them on what AI is, how it works and the implications it could have for the industry and organization by offering readings, speakers, courses and opportunities to attend events.
2. Make AI a priority and organize to win at it. The CEO could appoint a senior executive (not necessarily the chief technology officer/chief information officer) to be accountable for building and executing the AI plan and then communicate the plan to the organization to show commitment and begin funnelling ideas and talent to the AI lead. Hire AI entrepreneurs, researchers, technologists and data scientists/engineers to staff a centre of excellence (CoE). Like other disruptive technologies, AI works best in a "lab" environment that enables experimentation with technology and business applications.
3. Task the AI executives, business-unit leaders and CoE with identifying and sizing the organization's top AI-use cases – opportunities in which AI algorithms can be trained on internal or non-conventional data sources to create enormous value for the organization. The best AI-use cases capture large financial opportunities, lead to an advantage over an organization's competitors and/or significantly accelerate the organization's core business. This could involve enhancing a current product or service with, for example, better predictive abilities or consumer personalization; vastly improving an internal operation through an AI-enabled capability such as predictive maintenance; or even a creating a product or offering a service that doesn't exist today – such as a retail shipping service that delivers needed products to a consumer's house before he or she actually orders them, thus preventing purchases from competitors.
Use a test-and-learn approach as you begin to develop your selected use cases. It is important to experiment with different data, algorithms and approaches to predict various outputs rather than waiting for the perfect answer. Let AI talent work with business leaders to quickly launch, iterate and learn from AI solutions, and establish monthly deliverables to drive results.
4. As you develop your AI capabilities, it's also essential to stay connected to the AI startup scene. Some of the best use cases and talent will collect in these startups and watching them closely will lead to concrete opportunities. It will also ensure that you think about AI beyond immediate use cases and build awareness of how this platform technology can transform your industry and shift economic profits along or outside of the value chain.
In Canada, we are at the centre of AI and machine learning – much of the leading-edge science is happening here. But if Canada is going to lead the way in AI and reap the economic benefits, every business, government ministry and not-for-profit will need an AI action plan, just as they needed a web strategy at the dawn of the internet age. The boldest and fastest leaders and organizations will be best prepared to win.
As Steve Jurvetson, the legendary Silicon Valley entrepreneur and investor, said: "I believe that machine intelligence is the biggest advancement in engineering since the scientific method … most large companies don't realize how important it will be to their future."
Our challenge to leaders is to get in the game now – start building your AI action plan today.
Vincent Bérubé is a partner at McKinsey & Co. John Kelleher is a partner at McKinsey & Co. and co-chair of NEXT Canada. Tiff Macklem is dean of the Rotman School of Management at the University of Toronto.
On Oct. 26, 2017, Rotman will be hosting its third annual global conference on the business of AI – Machine Learning and the Market for Intelligence – where McKinsey & Company will be discussing its research on the implications of AI for various industries and occupations.