‘Hard, unglamorous work’ now can mean AI success later, city leaders say
Alex Potemkin via Getty Images
By
Chris Teale
Use cases for the technology are multiplying and now include areas like storm and wildfire prediction. But first, cities must make sure their data is clean and employees are on board.
NEW YORK CITY — When Hurricane Milton swept through Florida less than two weeks after Hurricane Helene made landfall, it also brought something officials had not seen before: tornadoes.
While it may be easy to dismiss that combination of events as out of the ordinary, it got some city leaders in Florida thinking about whether this could be the new normal as the effects of climate change become more acute. Those leaders are now turning to artificial intelligence to track weather patterns in a more detailed way and try to make such predictions.
“We had several tornadoes through Florida on the west side and across our city, and it was really just an anomaly,” Port St. Lucie Mayor Shannon Martin said during a panel discussion at the Smart City Expo USA conference in New York City last week. “But now we're starting to question, is this going to be what the future is with regard to the weather? If we have two more hurricanes that are within a week or two of each other because of all the atmospheric pressure, what's going on there?”
Port St. Lucie’s use of AI in emergency management comes as the technology gradually takes hold in state and city governments. Others are doing similar work: Kyle Patterson, the director of organizational effectiveness for Boise, Idaho, said his state is looking at using the technology to predict, prevent and respond to wildfires, so they can be “deploying the right resources in the right time, in the right place, to be effective.”
Those efforts in emergency management build on AI’s use cases in areas like process efficiency, automating workflows and finding information. Greg Useem, Alexandria, Virginia’s chief performance officer, said cities are at the “tip of the iceberg” of what is possible.
But while these use cases and others might be exciting, cities cannot just implement new tools and hope for the best. They must first do the “hard, unglamorous work” first, Patterson said, including making sure the data their AI models rely on is clean. Otherwise, it could result in incorrect results, hallucinations and other negative consequences.
That data will never be perfect, Useem said, but it at least must be cleaned and standardized enough to be used properly.
“Data will never be perfect,” he said. “We've got a ton of data in our organization. We're a full-service city. We've got things from the police department to health and human services and more. So the question is, how do we get data in good enough shape so that we're able to then go and leverage AI?”
Andre Sayegh, the mayor of Peterson, New Jersey, said during the panel that the COVID-19 pandemic provided the “lifelong lesson” that, when used properly, “data can help you save lives and make predictions,” as it did with dashboards showing infection rates, testing sites, vaccine availability and other features. Sayegh said the city has come a long way since he took office, when data “wasn’t used to benefit the residents of Paterson.”
Communication is key, too, as cities experiment with AI and look to get both their residents and employees on board. Public-sector employees appear to be less skeptical about the technology as they become more comfortable with it, and Patterson said there is clearly momentum. He talked up Boise’s AI Ambassador program, where one or two staff per agency use “peer-to-peer interaction” to encourage AI’s use among their colleagues where appropriate. Personalization is key, he said.
“I could send out a citywide email saying, ‘Generative AI is great, here's some use cases, here's some things to be careful of,’ and that's helpful,” Patterson said. “It's not useless, but it's not going to result in large-scale change. But if I'm a librarian, and my trusted colleague, a librarian says, ‘Oh, I've been using generative AI, there's these three ways that I've been using it, and you can use it also, watch out for this, you have to be careful,’ that's going to lead to real change in the way people work.”
While a lot of work lies ahead as cities look to embrace AI and its various use cases, speakers said the technology is not going to go away, even as progress is uneven. Sayegh said Paterson is “not where we want to be” on the technology but is making progress. Martin called AI the “way of the future,” while Patterson said money spent now will benefit generations of residents.
“This is a tool that we need to be able to thrive as a city, to provide our residents what they need, to be responsive to residents,” he said. “We all do that if we invest in it and we treat it as an asset.”