Less than five years ago, the artificial intelligence (AI) market was valued at a few hundred million dollars. By current estimates, the AI market is projected by analysts to reach somewhere between $30 billion and $50 billion by 2020. That’s just a couple of years away.

The rate of adoption in both product development and end-product applications has been stunning. It has been driven in large part by machine learning and natural language processing, which have created tangible business value. AI is set to power 85% of customer service interactions (Gartner) and in the next two years 40% of digital initiatives and 100% of IoT initiatives will all be AI based (IDC). In other words, AI will find its way into just about every product and service.

Customers don’t buy products; they buy experiences. The rapid success from the likes of Amazon, Google, Tesla and Waymo comes from a very practical application of AI. It is changing the way we think about delivering products and services, about automating systems and scaling expertise. It’s a new way to interface – and augment human.

Next AI frontier: intent and respect

Human-centred interactions between people and machines have profound implications on the design of products and services. No longer do consumers need to command machines using a graphical interface: voice interfaces such as Alexa, Siri and Cortana etc. have changed that. Next, the emphasis will shift from understanding the meaning to interpreting intent. For example, in Toyota’s Concept-i car instead of commanding its virtual AI assistant, Yui, to turn the AC up, Yui will be able to understand intent in statements like “I’m feeling a bit cold.”

>See also: Machine learning and AI: changing how data science is leveraged for digital transformation

It isn’t necessary to look into the future to see this trend. Already data-driven products are taking on board the emotional reactions of their users. For that reason, the best data-driven services don’t exhaust the user with endless data-gathering questions: Apple Music asks new users to “Tell us what you’re into” and presents a few bubbles containing genres to select. Netflix asks new users to select three movies they like at sign up; they handle the rest over time.

Similarly, data products need to know when not to suggest decisions for their users because ultimately, users want to remain in control. For instance, Nest Labs discovered that people don’t like being told what to do. For the Nest Thermostat, letting users feel in control led to a better experience and to increased energy efficiency. Nest’s initial Auto-Scheduler algorithm was optimised to reduce energy costs and because they failed to take the end user experience in to account, this algorithm led to higher energy usage. The Nest designers listened to their users and updated the Auto-Schedule algorithm to ensure comfort and respect user inputs.

Reading minds

We are all used to rudimentary predictive systems: predictive texting, personalised Netflix recommendations, Amazon’s shopping suggestions and so on. There’s a wide variety of applications, but they all share the same basic methodology: take a large quantity of historical user data, find correlations and then predict future needs.

AI technologies take predictive analytics to the next level. For instance, connected vehicles offer an opportunity to collect real-time user data on a massive scale. It’s the way self-driving cars of the future will be able to predict road conditions and prevent accidents. Another example is the way a pioneering Fortune 500 insurance company is currently using driving behaviour data collected from GPS and accelerometer smartphone data to offer timely, location-relevant promotional deals and financial advice, as well as encourage safe driving behaviour.

Context is king

While AI appears to have all the answers, the benefits are determined by the business case. Contextual knowledge brings the necessary focus to understand the business value proposition.

For example, an industrial-equipment maker will benefit from better predictions about the onset of failure so it can schedule maintenance in advance. Likewise, better predictions will help a mobile operator cut the time it takes to resolve network congestion and isolate instability. Alternatively, they can help a software developer speed up the release in the backlog of features by rooting out priority bugs much earlier. Indeed, a key use case for AI technologies is in optimising processes themselves, not just end products.

>See also: Put your money where your mouth is: How to make the most out of your machine learning investments

Dynamic data-based decision making

IDC predicts that three quarters of development teams will include AI functionality in applications or services. There is a need for professionals who add something extra beyond data literacy. They must be skilled at integrating machine learning into the core of products and services. Once the product launches, it’s important to continue to use the data available to validate the offering. Data-driven companies who spot trends are able to change their models accordingly. Forrester estimates that data-driven organisations will have a huge $1.2 trillion competitive advantage over their less data-savvy peers.

Interestingly, in organisations where AI is being used optimally, there is a much higher general understanding of intelligence and data-driven decision-making. In these organisations, information plays an integral role in the non-technical departments too.

What next? AI has the answer

AI is as much art as science. Success in AI demands a new set of skills among the development community and a commitment to developing production-ready, easy-to-use cognitive products and platforms. It requires a refreshed understanding of what makes product design great in the data-driven age. As the technology becomes more ubiquitous, businesses will need to gain a deeper, more nuanced understanding of its place in the design process. They will need to apply AI thoughtfully.

Sourced from Walid Negm, CTO, Aricent