Artificial intelligence is projected to contribute $15.7 trillion to the global economy by 2030 and Big Tech giants including Apple, Microsoft, and Google parent Alphabet are spending billions to secure their dominance.
And yet despite all the hype, nearly half of all companies remain firmly on the sidelines.
The adoption of AI has more than doubled in the five-year period through 2022, with capabilities added to a variety of functions ranging from robotic process automation, to natural-language text understanding, to facial recognition. But the proportion of companies using AI in at least one business area has been stagnant in recent years, and even declined from a peak at 58% in 2019 to 50% last year, according to survey data from McKinsey.
“Gen[erative] AI and some of these technologies are like swimming: You need to really swim in the water to learn,” says Ankur Agrawal, a partner at McKinsey who focuses on counseling health care leaders.
Agrawal advises C-suite leaders to bifurcate their approach to AI. “We tend to think of productivity as the biggest driver of return, but in many cases, it is the long-term strategic application and the ability to drive the long-term strategic direction of the company, which leads to outsize returns,” he explains.
For companies that are considering deploying AI for the first time, Agrawal says setting up a cross-functional team that aligns on the right business cases to leverage AI can be the greatest challenge. “It isn’t a technology problem, it is a business problem,” he says. From there, companies should set up the basic tech backbone needed to enable AI, including data infrastructure and cloud computing. After those two steps, he says, firms should establish use cases where they can find either near- or long-term value.
Boston Consulting Group advises clients to assess the business value of AI through a 10-20-70 formula. That means: 10% of effort to build machine-learning models, 20% involving high-quality data and technology implementation, and 70% focused on developing new business processes or transforming the way business functions operate.
Sylvain Duranton, managing director and senior partner at BCG, says companies exploring AI for the first time should start with the launch of a task force that includes IT and HR functions. They should then look at the products available on the market and plan for costs of deployment—because it will be expensive.
Business leaders should assess one or two functions where the mandate for AI use would be a full transformation with at least 50% in productivity gains, says Duranton. If it is less than that threshold, look at other areas of the business where the benefit is greater. One mistake companies make is that they don’t think big enough.
“You change the company deeply,” says Duranton, adding some aspects of a business will be eliminated while other opportunities are created. “You need to keep in mind what is the blueprint of what I’m building.”
“We have a very clear playbook,” says Intuit chief data officer Ashok Srivastava.
The business software giant starts by building strategies and then makes significant investments in AI and data that has a proven ability to deliver value to customers quickly. “It’s not just about AI,” says Srivastava. “It’s about the entire platform, the data analysis, and everything that goes together to scale this across the entire business.”
“The more I’ve gotten to know AI, I really think the value of AI accrues to who has the best data,” says Dr. Robert Blumofe, chief technology officer at Akamai.
Generative AI is unique among technologies in that it is very accessible. Just two months after launch, OpenAI’s ChatGPT reached 100 million users. Three out of four executives told McKinsey they had some exposure to generative AI either professionally or outside the office, and 22% use it regularly in their work.
But they remain skittish about deployment. Six out of 10 CEOs told consulting firm EY that “uncertainty around GenAI makes it challenging to develop and implement an AI strategy.”
“Business leaders are first and foremost a bit confused and I would say even a bit frustrated,” says Chris Perry, chief innovation officer at communications firm Weber Shandwick.
C-suite leaders are disoriented by the pace of change and how quickly generative AI models are developing. “It’s like the genie is out of the bottle and we’re trying to figure out what it means,” says Perry.
In the past year, Weber Shandwick has worked with dozens of clients to help orient them on how AI applies to their business and what implications they should be thinking about. After executives visit the firm’s virtual lab and work through strategic assessments, scenario planning, and messaging testing, Perry says the energy in the room shifts to excitement.
Because generative AI’s algorithms create text, images, and sounds, it creates an eeriness that some struggle to comprehend. “The feeling is more human-like because of the use of language,” explains Sharon Mandell, chief information officer at Juniper Networks.
“We believe humans and machines come together to be able to serve our customers better,” says Jim Fowler, chief technology officer at insurer Nationwide. With all processes for which Nationwide utilizes generative AI models today, there’s always a human in the loop to address potential errors.
Nationwide’s business unit leaders are expected to drive the digital strategy for their divisions, and that includes AI usage.
“If there is a major technology investment program underway, they should be baking generative AI into that program,” advises Will Bible, digital transformation and innovation leader of audit and assurance at consulting firm Deloitte.
Deloitte advises companies that are exploring AI to begin with a committee approach, but not only think about the legal and risk implications. Leaders should ask: How will AI flow into the business? “It’s not a one-person exercise,” says Chris Griffin, audit and assurance managing partner of transformation and technology at Deloitte.
There can be multiple teams or committees across an organization working through how and where to adopt AI. Diverse perspectives are needed to make sure a company gets it right, ranging from sales and operations to risk, compliance, and legal.
“Our clients and their boards are talking about this at every meeting today,” Griffin says. “And I suspect that pace will stay that way for some period of time.”
This article is part of Building Trust in AI.
