NEW YORK CITY — Despite some existential dread about artificial intelligence (AI), Serge Leontiev thinks the technology is pretty safe for the time being. In fact, the public sector innovation lead at Oracle thinks humanity is at the monkey-smashing-things-with-a-stick stage of AI.

“At this point we are just starting our evolution and we’re trying to find out more what the technology can deliver,” Leontiev said.

But at the New York City Technology Forum, Leontiev said government is figuring out some early ways to pull value out of AI. Namely, it can allow government to pull value out of its troves of data, it can allow for better service delivery and it might even be able to replace the functionality of apps.

At the state and local government level, many IT officials and vendors have started working with AI as a way to power chatbots — conversational user interfaces that can automate specific customer service functions and help citizens navigate the government architecture.

An example: Many cities have mobile applications where users can do various things like access transit information or apply for services. But not everybody will be familiar with that city’s workings, or have the app installed.

“I’m not going to install a mobile application for each and every city that I’m visiting,” Leontiev said at the Tech Forum. “I’m not going to install a mobile application for each and every airport I’m passing by. But I would love to have a way to communicate with the government, report dirty platforms or get updates on traffic, or get information on services nearby.”

A chatbot set up the right way can get around those issues. If a person wants to report graffiti, for example, they might be able to do so using what they already have on their smartphone, rather than downloading an app or going to a website. Facebook, for example, allows for chatbots that can help governments interact with citizens.

“With Facebook messenger I can use all my device features, and I can snap a picture of the graffiti with my phone, and I can upload my geolocation,” he said. “So, at this point I’ve given enough information for the maintenance team to remove that graffiti.”

It might not be so simple for everything a government does. In many places, applying for a business license means going to several agencies within the city, handing over the same information time and time again, and then repeating at the county and state levels.

A chatbot can help guide a person through that process in a way more customized to the individual.

“It will ask you all these questions, and based on your answers will ask additional questions to create an outcome that will match with your actual requirements,” Leontiev said. “It will provide links to all these agencies that you have to contact and get information that will help you open a business license.”

Leontiev thinks AI might also help government become more proactive with helping citizens. Though government agencies hold a lot of data about individual people, they don’t necessarily use that data for what it’s worth.

“They’re just sitting on piles of data and hoping to do some business analytics and stuff, but they’re not thinking from the different point of view that you can use that data and machine learning techniques to figure out personal preferences for this particular individual,” he said.

It’s not just in the government that Leontiev thinks AI is at an early stage — it’s everywhere. It’s at a low adoption rate because people are just starting to figure out the potential of the technology.

But if one wants to get a glimpse of what AI looks like when it’s widely used across the public and private sectors, Leontiev has some advice: Look east.

“It’s very interesting, if you take a look at the Asian market, they are using chatbots every single day for everything from banking to city services,” he said. “We are in the very beginning stages of using that… they are reaching out so that an application can connect with any agency.”