AI’s Broad Effect

In terms of public safety, AI advances could dramatically improve incident response times.

This month’s issue features an excellent article by Phillipa Maister about artificial intelligence (AI). It’s amazing how quickly this technology has embedded itself into our daily lives. It has crept up a bit in the sense that we often don’t know when we are using it.

Businesses are using it to thrive, offset the worker shortage and improve customer experience. But how could AI affect such areas as public services, zoning and infrastructure? That’s a question that the Atlanta Regional Commission (ARC) has been working to address for the last three years.

Georgia Ai

Photo Gerd Altmann

The quasi-governmental agency covers 11 counties in Metro Atlanta and helps distribute federal funding for infrastructure and transportation needs. I discussed the potential impact of AI with ARC Executive Director and CEO Anna Roach and former Board Chair Sam Olens, who was also attorney general for six years and is now a partner in Denton’s public policy practice.

One of the first things Roach mentioned was the ability for AI to mitigate traffic – to which I say, bring it. “In the not-too-distant future, AI-powered technology will be used to analyze … data from cameras and sensors along our roadways to pinpoint danger spots, predict where accidents may occur and enable proactive measures to prevent collisions. Traffic signals could be tightly coordinated and managed in real time to respond to real-world conditions. More sophisticated data analysis could enable transit systems to better link transit service to job centers and health care providers,” she says.

While ARC is currently working to install 300 to 400 electric vehicle charging ports, autonomous vehicles (AV) could soon be on the agenda. “AI may be used to create AV systems that not only drive themselves but, when connected, continuously ‘talk’ to each other and travel in closely coordinated fashion,” Roach says. “This would enable vehicles to travel more closely together in the same amount of space and dramatically reduce crashes.”

There are big implications for land use. “Commercial buildings may no longer need parking lots or decks, making it easier to create more walkable, vibrant places. Without the need to provide … parking, building costs for single- and multifamily housing could drop sharply, helping ease our affordability crisis. Existing garages and parking lots could be turned into housing or workspaces,” she says.

The price of accommodating parking requirements, especially for multifamily development, increases the total housing cost by more than $50,000 annually per unit, she says.

“The ARC is a convener, not decision maker for local governments,” notes Olens. “But if you look at Gwinnett and Cobb, who for the first time are looking at revisions to zoning [and parking] codes, you see that much of that energy is on ordinances that candidly don’t serve adequate interest. Many developments are built with excessive parking, and now is a good time to look at revision.”

In terms of public safety, AI advances could dramatically improve incident response times. Researchers are exploring the use of AI-powered drones that can quickly deliver lifesaving medicine and equipment, such as defibrillators, that people can use while awaiting an ambulance, says Roach. “This could be especially useful in parts of our region that are long distances from the nearest hospital and now endure long waits of up to an hour for ambulance service.”

AI could also help respond to disasters. AI-powered cameras or satellite images can quickly evaluate damage and direct the response accordingly.

AI can even help deter crime. The Knightscope K5 Autonomous Security Robot, which uses cameras and other sensors to detect suspicious behavior, has been used in airports, casinos, hospitals and even a New York subway station.

Speaking to the dangers of runaway Robocops, Roach and Olens note “inherent biases and [AI] hallucinations” and the need for human oversight of AI. When I mentioned that on the internet (unlike the printed page) there are no consequences for false information that might lead AI astray, Olens said “you are 100% correct – not even AI can stop bad actors, so our role is to work with local governments to mitigate the effect. You can’t remove that risk; you can only minimize it.”

Advanced manufacturing and robotics still require a human touch. “No one should be relying 100% on AI, since you can’t code away vulnerabilities of the tool,” says Roach. “AI can take into account flaws, but human review and intervention will be a critical step we ought not forget.”

The Atlanta Regional Commission hosts an Emerging Technology Roundtable quarterly to discuss AI and other issues.

Categories: From the Publisher, Opinions