When people talk about AI, Jason Harper says they often focus on "autonomous vehicles or image recognition or other things that would be bleeding-edge AI technology research."
Harper, founder of the Ann Arbor cloud-based software company RXA, says those discussions are "great and exciting." But they overlook the many smaller problems those in the AI field are solving – and advancements they're making – right now. That's why RXA and Utah-based Domo created a one-day conference called A2.AI, held at SPARK Central in downtown Ann Arbor on Oct. 25.
A2.AI brought together AI experts and business managers to talk about the interaction between humans and AI, and how to use both to decrease operational expenses and improve efficiency. The event was cosponsored by numerous local software companies, Ann Arbor SPARK, and DTE.
Speakers included Don Barnes III, president of Belle Tire; Tim Freeth, head of industry for Google; and Douglas Kramon and Jennifer Lien from ESPN's fan support department. The day ended with a panel discussion moderated by Will Thompson, managing director of Forbes Insights and publisher of Forbes AI.
Don Barnes III speaks at A2.AI.
Harper says the vision for the day was to have business managers talking about practical applications and how they were using current AI technology to make decisions. Michigan-based Belle Tire, for example, is saving money on labor and reducing operational expenses using a workforce optimization app.
"It's about helping business managers understand what's available from an applied AI perspective now, that they can use today," Harper says.
"Don't forget the human intelligence"
A2.AI's first keynote speaker was Ben Schein, vice president of data curiosity at Domo, setting the conference's theme with a talk entitled "Don't forget the human intelligence on your AI journey."
Schein worked in Target's finance department for many years and was later recruited by Domo, whose staff created his current position just for him.
Schein said that "data curiosity is about how you change the culture," and how companies can encourage employees to engage with and explore findings from artificial intelligence and machine learning.
Ben Schein speaks at A2.AI.
He said when people are discouraged from asking questions about how AI and machine learning work, then AI becomes "this black box that no one understands" and people become scared for their jobs, rather than being excited about AI's potential.
"What I usually see happening is that a company invests a lot of money in AI and machine learning … and then waits for magic," Schein said. "But often the magic is not that impressive for the business person. It's just an output file."
The solution to that, Schein said, is to "honor the human intelligence" by asking for and listening to feedback from experienced employees who can make those AI algorithms even more powerful and nuanced.
One way to keep workers from disengaging is to "show the work" and refrain from over-summarizing. Employees don't need to dig into all the math involved in the algorithm, but it's helpful to show them some of the thought process that went into creating the algorithm.
Schein said "over-summarizing" has problems as well, because then employees can't engage and give feedback in a meaningful way. AI and machine learning programs must be built in such a way that they can accept changes, fine-tuning, and new contexts.
"You need to include flexibility to handle different business contexts, like a new line of business or a new competitor or having a competitor exit the scene," Schein says. "You can't build something that's not flexible when priorities are changing all around me."
Chris Curry, president of Tennessee-based Evaporcool, also noted the importance of human expertise in AI. Evaporcool's technology helps increase the efficiency of air conditioning equipment through the power of evaporation. But at certain times of day when the sun hits automated sensors directly, a false high reading triggers the cooling system to kick on when it’s not really needed.
Chris Curry speaks at A2.AI.
"Engineers, using their own internal expertise, can create a data capsule that tells the system to ignore that phenomenon," Curry said. "A pure machine learning approach won't achieve the value you need. You need the expertise of the engineer, who really understands what's going on out there in the field."
The RXA team felt now was the time for an AI conference in Ann Arbor because they couldn't find exactly what they wanted at existing AI events, which are typically very high-level and tech-heavy, Harper says.
"I used to be a data scientist, and I love coding and getting in there. But in reality, now I really just want to understand what people are doing with it, not how they're doing it," Harper says. "We wanted to talk about what's being done and get excited and think about how we can implement those things. That's what this conference is meant to be, an approachable discussion of AI. I haven't seen that out there."
Jason Harper speaks at A2.AI.
Brandon Rea, RXA's chief revenue officer, says RXA knew it had "the know-how, the ingenuity, the expertise, and the contacts" and wanted to bring that all together in one place so participants could share what they've been working on.
"We think Ann Arbor is the perfect place for this, certainly because of the University of Michigan, but also because healthcare meets technology and the automobile industry," Rea says.
The event turned out to be more popular than organizers expected and was at capacity two weeks before registration was set to close. Harper says he expects to host more AI conferences in Ann Arbor, at least annually, and maybe more often depending on demand.
Sarah Rigg is a freelance writer and editor in Ypsilanti Township and the project manager of On the Ground Ypsilanti. She joined Concentrate as a news writer in early 2017 and is an occasional contributor to other Issue Media Group publications. You may reach her at firstname.lastname@example.org.
All photos by David Lewinski.