‘Aware of AI’s potential but unsure how to seize it’ — that’s where many companies are today. But some are pushing ahead, building small blocks of value into a transformative whole.
PwC’s 2019 AI Predictions report surveyed 1000 executives at US companies that are currently investigating or implementing artificial intelligence initiatives. Of these AI leaders, one-fifth plan to take the technology organisation-wide in the coming year.
That may seem ambitious, but if done right, these plans will pay off.
Forty-eight percent of survey respondents expect AI initiatives to grow revenue and profits in 2019. Forty-six percent expect better customer experiences. And 40 percent expect AI to improve human decision making.
It’s a good thing that the rewards are so great, because the obstacles are too. How to trust complex algorithms? How to get employees who aren’t computer scientists working with AI? How to get AI to turn data into value?
And then come the big questions one asks about any new technology: how do we actually roll it out, and how do we make money with it?
The survey data, alongside experience, offers the answers by highlighting six priority areas businesses should address.
The first priority is based on a fact about AI algorithms that may surprise business users: There aren’t that many of them. The same few algorithms can solve most business problems for which AI is relevant, so if you successfully get them producing ROI in one small part of your business, you can usually use them in others.
That makes it possible to turn localised ROI into enterprise-wide momentum, if you have the right structure: a team of people with business, IT, and specialised AI skills who represent all parts of your organisation. An AI centre of excellence (CoE) is often the best approach — and the model we expect to be most prevalent. (We have an AI CoE at PwC.)
AI is hard. Even with basic training, business people may not fully understand different algorithms’ parameters and performance levels. They could accidentally apply the wrong algorithms, with grave results.
That’s why 36 percent of survey respondents called upskilling for AI a top challenge for 2019. Twenty-eight percent cited recruiting challenges.
Companies have to prioritise upskilling and recruiting to build the right mix of employees: some who just know how to use artificial intelligence applications (citizen users); others who can work with AI specialists to create new applications (citizen developers); and the specialists themselves.
You’ll then have to give these employees the tools and skills to work together, while fostering the culture that both attracts top talent and lets them do great work.
Hold on, you may say. You want me to put our cyber defense, marketing strategy, customer experience, and more in the hands of software so fast and complex, even computer science PhDs aren’t always sure how it works?
The answer to that question is yes, but it’s not surprising that 37 percent of survey respondents deemed AI’s trustworthiness a top challenge.
To make AI trustworthy, prioritise the five dimensions of responsible AI: fairness, interpretability, robustness and security, governance, and systems ethics. That’s possible with proven best practices for controls over emerging technologies, and with the recent advances in explainable AI, which can tell you its rationale, strengths, and weaknesses
How to turn data into value? Add AI. The top AI-related data priority for 2019 (cited by 58 percent of respondents) is to integrate AI and analytics systems to gain business insights.
Show AI enough historical data on consumer behavior, for example, and it can eventually predict how those consumers — and others like them — will behave going forward.
But for data to teach AI, you have to first locate the right data, then label it: whether a data point shows a satisfied or unsatisfied customer, to take a highly simplified example. New AI tools can help: lean and augmented data learning techniques, for example, can enable AI to produce its own data based on just a few samples.
AI’s greatest value today is from productivity enhancements, as businesses use AI to automate processes and help employees make better decisions. But as our global artificial intelligence study, Sizing the Prize, found, the majority of AI’s economic impact will come from the consumption side, through higher-quality, more personalised, and more data-driven products and services.
So it’s time to prioritise the new business models and strategies that AI makes possible. Retailers, for example, are using artificial intelligence to anticipate trends and guide the business to meet them, while a leading auto manufacturer has been using AI to test more than 200,000 go-to-market scenarios for autonomous ridesharing fleets.
It may feel like juggling champagne glasses while riding a unicycle into the wind, but companies have to do it: combine AI with other still-evolving technologies. It’s the only way that, for example, an enterprise will be able to manage millions of Internet of Things sensors — and monetise the data that they produce. Thirty-six percent of the executives surveyed called the convergence of AI with other technologies a top challenge for 2019.
DevOps techniques, which put development and operational teams in a feedback loop for constant collaboration and interactive changes to new products, will support a seamless convergence. So will new roles for employees to serve as translators and liaisons among the various technology teams.
AI holds enormous potential. It can enhance cybersecurity and the customer experience, cut costs and improve forecasts, innovate new products, and create new business models and revenue streams. Since companies that advance quickest with AI will have access to ever more valuable data, the first-mover advantage could be immense.
With payoffs possible right now, and a competitive edge likely for the future, every company should make AI a priority today.
For further detail on the areas businesses must prioritise to make the most of artificial intelligence, visit the full PwC’s 2019 AI Predictions report.
© 2017 - Thu May 26 16:18:20 UTC 2022 PwC. All rights reserved. PwC refers to the PwC network and/or one or more of its member firms, each of which is a separate legal entity. Please see www.pwc.com/structure for further details. Liability limited by a scheme approved under Professional Standards Legislation.