Among the many business disruptions caused by covid-19, here’s one largely overlooked: the whiplash of artificial intelligence (AI).
As the pandemic began to rock the world last year, companies have used every tool at their disposal, including AI, to solve problems and serve customers safely and efficiently. In a KPMG 2021 survey of U.S. business executives conducted between Jan. 3 and Jan. 16, half of those surveyed said their organization had ramped up its use of AI in response to covid-19, including 72% of industrial manufacturers, 57% of technology companies and 53% of retailers.
Most are happy with the results. Eighty-two percent of those surveyed agree that AI has been useful to their organization during the pandemic, and a majority say it offers even more value than expected. More generally, almost all say that wider use of AI would make their organization more efficient. In fact, 85% want their organization to accelerate the adoption of AI.
Still, the sentiment is not entirely positive. Even though they are looking to press the gas, 44% of executives believe their industry is moving faster than it should on AI. More surprisingly, 74% say that using AI to help businesses remains more fashionable than reality, up sharply in key industries since our September 2019 AI survey. In service industries finance and retail, for example, 75% of executives now believe AI is overrated, up from 42% and 64% respectively.
How do you reconcile these seemingly opposing views on what KPMG calls the whiplash of AI? Based on our work helping organizations apply AI, we see several explanations for the hype. One is the sheer novelty of technology, which has allowed for misperceptions about what it can and cannot do, how long it takes to achieve enterprise-wide results, and what mistakes are possible when organizations experiment with AI without the right foundation.
Even though 79% of respondents say AI is at least moderately functional in their organization, only 43% say it is fully functional at scale. It’s still common to find people who see AI as something to buy, like a new machine, to get immediate results. And while they’ve had some success with AI (often small proofs of concept), many organizations have learned that moving them to the enterprise level can be more difficult. It requires access to clean and well-organized data; a robust data storage infrastructure; subject matter experts to help create tagged training data; sophisticated computer skills; and company buy-in.
Of course, it’s also no exaggeration to believe that supporters of AI may have exaggerated its potential every now and then or reduced the effort required to realize its full value.
As to why leaders are in conflict over the speed of adoption of AI, we see fundamental human nature at play. For starters, it’s always easier to believe that the grass is greener on the other. side. We also suspect that many people are concerned that their industry is changing too fast, mainly because their own organization is not keeping up with that speed. While they experienced early stage hiccups with AI, especially last year, when the world witnessed achievements through AI, like the record development of covid-19 vaccines, it Perhaps it was easy to succumb to these fears.
We see another factor that elicits mixed feelings about AI’s potential: the lack of an established legal and regulatory framework to guide its use. Many business leaders don’t have a clear vision of what their organization is doing to govern AI, or what new government regulations may be on the way. Understandably, they’re worried about the associated risks, including the development of use cases today that regulators could crush tomorrow.
This uncertainty helps to explain another apparently contradictory result of our investigation. While business leaders are generally skeptical of government regulation, 87% say government should play a role in regulating AI technology.
Go from whiplash to AI
While every organization will need its own manual to recover from the AI whiplash and maximize its investment in technology, a comprehensive plan should include five elements:
- A strategic investment in data. Data is the raw material of AI and the connective tissue of a digital organization. Organizations need clean, machine-digestible, labeled data to train AI models, with the help of subject matter experts. They require a data storage infrastructure that transcends functional silos within the enterprise and can deliver data quickly and reliably. Once the models are deployed, a data collection strategy and approach is needed to tune and train them on an ongoing basis.
- The right talent. IT people with AI expertise are in high demand and hard to find, but crucial to understanding the AI landscape and guiding strategy. Organizations unable to assemble a full team of in-house scientists will need external partners who can fill in the gaps and help them sort through the ever-growing number of AI vendors and offerings.
- A long-term, business-driven AI strategy. Organizations get the most out of AI by thinking about finding solutions to problems, not buying technology and looking for ways to use it. They let the business, not IT, run the agenda. When AI investments tied to a business-led strategy go awry, they become opportunities for quick failure and learning, not quickness and scorching. But even if companies iterate quickly, they must do so in accordance with a long-term AI strategy, because the greatest benefits are realized over the long term.
- Culture and skills development of employees. Few AI programs will gain traction without workforce buy-in and a culture invested in AI success. Gaining employee engagement requires providing them with at least a rudimentary understanding of technology and data, and an even deeper understanding of how it will benefit them and the business. It is also important to hone the skills of the workforce, especially when AI takes over or complements their existing responsibilities. Adopting a data-driven mindset and instilling a deep knowledge of AI into an organization’s DNA will help them grow and succeed.
- A commitment to the ethical and impartial use of AI. AI holds great promise, but also potentially damaging if organizations use it in ways that customers dislike or discriminate against certain segments of the population. Every organization should develop an AI ethics policy with clear guidelines for how the technology will be deployed. This policy should enforce metrics and be part of the DevOps process to check for problems and imbalances in data, measure and quantify unintentional biases in machine learning algorithms, track where data comes from, and identify those who shape the algorithms. Organizations must continuously monitor models for bias and drift, and ensure that the explainability of model decisions is in place.
Executives’ goals for AI investments over the next two years vary by industry. Health officials say they will focus on telemedicine, robotic tasks and providing patient care. In life sciences, they say they will look to deploy AI to identify new revenue opportunities, reduce administrative costs and analyze patient data. And government leaders say they will focus on improving the ability to automate and analyze processes, and manage contracts and other obligations.
Expected results also vary by industry. Retail executives predict the biggest impact in the areas of customer intelligence, inventory management, and customer service chatbots. Industrial manufacturers see it in the design, development and engineering of products; maintenance operations; and production activities. And financial services companies expect to improve in fraud detection and prevention, risk management and process automation.
Over the long term, KPMG sees AI as playing a critical role in reducing fraud, waste and abuse, and helping businesses refine their sales, marketing and customer service operations. Ultimately, we believe AI will help solve fundamental human challenges in areas as diverse as disease identification and treatment, global agriculture and hunger, and climate change.
It’s a future that deserves to be worked on. We believe government and industry have a role to play in making this happen – working together to formulate rules that promote the ethical evolution of AI without stifling the innovation and momentum already underway.
Find out more in the KPMG Thrive in an AI World Report.
This content was produced by KPMG. It was not written by the editorial team of the MIT Technology Review.