There’s nothing like a technological paradigm shift to test your organization’s health and ability to absorb change. Just like businesses had to embrace mobility and cloud computing or risk being left behind, they’re now having to find ways to use AI effectively. This time, though, the stakes are even higher.
The initial rollout of OpenAI’s gen AI chatbot ChatGPT in late 2022 was the starting gun for a technology race that’s seen endless LLMs compete to gain attention of digital and business leaders. So senior executives must find technologies to help their organizations gain a winning advantage from a glut of models and AI-enabled vendor services. Research suggests establishing this edge is no easy task.
Many organizations have launched dozens of AI PoC projects only to see a huge percentage fail, partly because CIOs don’t know whether they meet key metrics, according research from IDC. Lily Haake, associate partner at executive search and leadership consulting specialist Savannah Group, says evidence shows a business that’s good at delivering transformation will probably be good at wielding AI.
“A significant factor for success is business change,” she says. “AI can’t exist in a vacuum since its effectiveness depends on the workforce’s ability to utilize it effectively. Our clients with a tried-and-tested blueprint for implementing business change tend to achieve better commercial outcomes from AI implementations.”
Putting processes first
One digital leader who’s helped her organization embrace AI-enabled transformation is Wendy Redshaw, CDIO at NatWest Retail Bank. She’s led the introduction of Cora+, NatWest’s next-generation assistant powered by gen AI. This multichannel agent, which first went live last June, provides natural answers to customers using data from multiple sources, including products, services, and banking information.
Redshaw looks back on the project’s inception and recognizes it was a good barometer of the bank’s ability to absorb and embrace change. “The driver was that gen AI is coming and how can we exploit this technology so we can stand behind it and say we’re happy,” she says. “And then we started thinking how can we do this work safely.”
The good news for NatWest is the organization was in a healthy position to answer these questions. The bank used its existing AI and data principles, including confirming that emerging technology is subject to human oversight, respectful of human agency, and technically robust, resilient, and safe.
“We were mindful of preventing unintended harm to individuals,” she says. “Cora+ had to be free from unfair bias, so we needed to be crisp and clear on the data, and make sure it wasn’t discriminatory. And we considered the social and economic impact because gen AI runs the risk of being very consumptive of energy.”
Redshaw’s common-sense approach to assessing data-led projects is rarer than expected. Data consultancy Carruthers and Jackson recently released its Data Maturity Index, which suggests nearly 40% of organizations have little or no governance framework. CEO Caroline Carruthers says executives must address these persistent gaps in foundational practices by prioritizing human agency.
“There are so many nuances data and AI don’t understand, but a human does,” she says. “What we should do is use data and AI to help us make better decisions, feed our curiosity, and help us be more innovative. That’s how we should use it, rather than say, ‘Computer says X, so we have to do X.’”
In the case of NatWest placing human agency at the heart of its AI processes, the bank ensures every model follows the same form of governance. For a pioneering project like Cora+, the leadership team considers AI ethics and principles at the outset.
Redshaw says executives sit together in a room and consider potential implications as parties with a vested interest. While she might rigorously scrutinize the environmental impact of AI, other executives act as regulators or customers in vulnerable situations. She says this kind of collaborative tactic means every element is considered from the start of a project. And it’s an approach that continues during rollout.
“There’s now an official way of doing things, and everything has to follow that path,” she says. “But in the very early stages, where you’re using a nascent technology and doing something different, we make my team work in an agile way, so they tend to think about how to do this work.”
Prioritizing personal requirements
That sentiment resonates with Zakir Mohammed, manager of AI and automation at Toyota North America. He says AI is here to stay and the key role of digital leaders now is to help their businesses embrace these advances effectively.
“We should be positive,” he says. “Using these technologies, we can improve productivity and give opportunities to our team members to do more of what they love to do. If I put a person in an eight-hour job working on spreadsheets or installing software, I don’t think that’s sustainable. We need to unleash their potential to come up with new ideas.”
Then there’s Adobe CIO Cindy Stoddard, who’s led her company through multiple tech-led transformations during the past decade. She also recognizes the importance of AI to the future of work and is busy using her experiences to create a template for change. “We want our staff to have an AI helper by their side assisting them do their strategic work,” she says.
The company created specialist group AI at Adobe last year to help shape its strategy, and includes people from legal, engineering, IT architecture, and other business functions, and its representatives set the direction of travel for emerging technology.
“This group is part of the governance process we have around AI,” she says. “If we introduce AI-based software solutions or develop applications internally, we do it correctly. Our product and engineering teams also go through this process to ensure we run our developments safely and with the customer in mind.”
The key part of this governance process is personas, finding the right tool for the right individual. Stoddard says Adobe doesn’t want to end up with gen AI sprawl, where the company has a complex tangle of services. Instead, AI at Adobe works with the IT team to control the organization’s AI tools, mainly Microsoft and in-house developed services.
She adds how Adobe uses AI services to support staff. The company’s finance division uses emerging technologies for forecasting and analysis, and its treasury division uses AI to investigate customer payment requests. In her own IT department, Adobe has discovered a range of persona-based use cases.
“If I look at program managers, for example, they have to read a lot of documents, go to a lot of meetings, look for risks, and things like that,” she says. “We’ve put together some gen AI capabilities that find particular action items that were opened but haven’t been followed up on yet. Our approach is about giving people tools to focus on strategic things rather than look for what they need to do.”
Making time for education
Stoddard says the key lesson from Adobe’s nascent explorations into AI is that an organization will be best placed to take advantage of emerging technology by tightly tying use cases to work personas.
“That’s how you get the value out of the technology,” she says. “Not by just throwing tools out and saying you may get a bit of value on your desktop, but you won’t get the same value as if you’re collaborating with AI and using it as part of the working process.”
In short, organizations in the best health to maximize AI pay close attention to staff requirements. However, research indicates much work needs to be done in this area. While 53% of digital leaders responding to Carruthers and Jackson’s Data Maturity Index reported increased AI use in their organizations during the past 12 months, over half said most employees lack data literacy.
Carruthers says the lack of progress is concerning as most CIOs recognized the requirement for increased data awareness 10 years ago. “I want to see more people understand how AI can improve their working lives,” she says. “Digital leaders must think about helping staff understand data better and how that knowledge will feed into their organizations.”
LinkedIn chief product officer Tomer Cohen also points to the importance of training and development. Businesses that absorb and embrace change most effectively dedicate time and resources to cultural processes, and ensure staff are confident using AI in their roles.
“A great, adaptive, resilient organization can transform their employee base,” he says. “And part of that success is giving people time to learn. If you only expect your employees to perform and do their regular tasks, they won’t have the space to learn new tools.”
Cohen refers to an example from his organization, the introduction of Full Stack Builders (FSB) last November. FSB uses AI and other emerging tech to shift the emphasis of product development at LinkedIn. Rather than a series of discrete tasks, such as designing and engineering, product development is undertaken by one professional assisted by AI services, such as tools that deal with coding assistance and product management.
“It requires time to learn these new tools,” he says. “You can say, ‘Once you learn these new tools, things will be easy for you,’ but there’s a learning curve. So, you must think about how you enable your employees to spend time on that learning curve. Because if you don’t, you’re just asking them to perform all the time at the expense of learning.”
Leave a Reply