Humans drive AI.
Despite its popularity, artificial intelligence has the problem that people don't trust it. This is a strange and complex endeavor for many, with terms like “diffusion models” and “emerging behavior” being thrown around.
Add to that the illusions, biased results, and general fear that AI will take away jobs, and AI expanding across organizations is definitely unwelcome. On the contrary, there is a tendency to rely too much on the output of AI, which can lead to all sorts of complications.
There is a middle ground somewhere. It's about being cautious, staying close to the business, and embracing all that it has to offer.
“When you think about generative AI, you have to think about the role it plays,” said Karen Feldman, vice president of marketing and communications at IBM. Feldman led a panel discussion at the recent Adobe Summit that looked at trust issues in AI and how to address them. Going forward, she said, generative AI will shape “the trust that we need to build with our brands, the trust that we have with our customers, and the trust that we need to build with our own employees.” “The only way to take advantage of this magical technology is for employees to embrace it themselves.”
Furthermore, she continued. “I hear questions about AI from my clients: “Can I trust its output?'' “Will it resonate with my customers?'' “Is the AI I'm using safe?'' “Is it brand compliant?'' “Can we trust that AI will reward individuals for their creations?”
The key to a successful AI business is making sure everyone in the room is comfortable using AI. For Aramark, one of the largest service companies, that means establishing a “safe space” for discussion and experimentation, said panel participant Laura, Aramark's vice president of B2B marketing operations and communications.・Mr. Powers said. “Trust in AI is very negative right now,” she said. “It's like, 'We're all going to die because of AI.'”
Powers aims to make AI easy to understand, sustainable, and open to all employees with ideas about how AI can improve their work. She said her division is driving and piloting AI initiatives and serves as an internal “authority” for the rest of Aramark.
Aramark's AI efforts began with the creation of an AI task force, she said. “I know it sounds very big brotherly, but it was actually reassuring to have AI as something that we knew we were starting to talk about in public,” she said.
The main focus of the task force's work was to develop comprehensive policies as part of governance. She explained that the policy has two main points.
- “Never input client or proprietary data into an AI tool.”
- “AI is not a replacement for human judgment.”
With these two guidelines, “we really began our journey at Aramark with the ability to feel comfortable in the AI space,” she said.
The company's AI Task Force is comprised of two committees: the Regulatory Committee and the Investigation Committee. “A regulatory committee comprised of our compliance officer and legal department oversaw the management of risk, the assessment and definition of risk, the determination of risk levels and the establishment of governance,” Powers said. “When an idea moves forward, regulatory commissions look at the technology behind it. Is it a closed environment? Does it contain Aramark materials? Is it safe for us?”
The AI review committee is “comprised of many departmental and cross-functional ambassadors from around Aramark,” Powers said. “We talked to groups of people about AI and what they’re interested in, what they’re afraid of, what they’ve heard, what they’re researching, what they’re doing. We can talk about it in a comfortable and open way.”
AI “ambassadors” are located throughout the company to provide support and listen to employee feedback as AI projects get off the ground. We are actively soliciting ideas from our employees. Provides an online form for submitting ideas for AI projects. “You can submit that form, or you can talk to the closest AI ambassador in your department or any ambassador to submit it for you.”
Aramark also speaks regularly with technology partners such as IBM, Salesforce, and Microsoft to “discuss what they're doing with AI within their companies, which has been an eye-opener.” said Powers.
Another channel for communicating employee ideas is a team channel called AI Buzz and a companion channel called AI Pathfinders that provides project updates. “This is a continuous, cyclical process that we're doing, and we're here to encourage this safe space for AI exploration,” Powers said.
An early AI application the task force is pursuing is proposal generation, “creating an initial draft proposal from an RFP or RFQ,” Powers said. “We started looking at this proof of concept, called Intelligent Response Assistance, after we identified these kinds of discrete areas where we thought we could improve our efforts. We want to give agency team members, salespeople and subject matter experts across Aramark more time to draft proposals.”
The team will apply AI to measure “time spent on proposals, money spent with agencies, and quality of proposals.”
Other projects implementing AI include legal chatbots and human resources automated virtual assistants, Powers said.
Powers offers the following advice to keep your AI efforts on track, but stay grounded.
Build AI based on your business case. “If you're interested in getting started, you don't know where to start, or you're in the middle of a journey, a business case is really important,” says Powers. “We did a lot of work thinking about value drivers, business objectives, what are we measuring, what is the business case, and how do we make this investment really worthwhile? .”
Start small. First of all, “we focused on just one line of business,” she said. “The department partnered with us for a proof of concept. They subscribed and attended all the discovery sessions. The demo came together really quickly.”
Think about skills. Let's ask some bigger questions. “What skills do I need to develop with my team members?” Powers recommends. “'What kind of organizations do we need to work with?' 'Do we need to reorganize?' I think the answer is clearly yes. ”
Establish shared governance. “What's really important in establishing governance is that all departments are involved,” Powers said. “To get this started and to define how these committees work and set up the channels, we needed to have a shared voice. Many of us felt that we had a role in governance as well. I felt that there was.”
Please stay positive. “Have a conversation internally about what your colleagues are thinking,” says Powers. “Find evangelists who can help you and want to talk about AI together. I found a like-minded group, and those people really help you stay on a positive and upward trajectory.” He will help you.”
