From ChatGPT to Tesla’s Autopilot, Mason professor Missy Cummings isn’t afraid to call out bad tech

Melanie Balog — Fri, 03 Feb 2023 19:01:27 +0000

From ChatGPT to Tesla’s Autopilot, Mason professor Missy Cummings isn’t afraid to call out bad tech Melanie Balog Fri, 02/03/2023 - 14:01

In This Story

Missy Cummings

Missy Cummings.
Photo by Ron Aira/Creative Services

Missy Cummings made her name as one of the United States’ first female fighter pilots. But it is also in her current career researching artificial intelligence, human-autonomous system collaboration, and the socio-ethical impact of technology, that she has made headlines.

Cummings, a George Mason University professor in the computer science, electrical and computer engineering, and mechanical engineering departments of the College of Engineering and Computing, calls herself a “tech futurist,” whose job is to “make tech work. It’s not to stop tech, it’s to help it get better.”

One of the ways she does that is to call it as she sees it, such as her take on ChatGPT.

“If you really pay attention, you can see very quickly how wrong and dangerous disinformation coming from something like ChatGPT could be,” she says. “Intelligence technologies are advancing so rapidly, what we’re not doing is keeping up with allowing people to get educated in how to think about the design frameworks behind when you should have these systems. Why should you have these systems? What requirements are they really meeting? And, then, how should I test these systems to make sure they are sufficient?”

As a safety advisor from 2021 to 2023 for the National Highway Traffic Safety Administration (NHTSA), she also has a good grasp on the evolution of self-driving cars, particularly when it comes to Teslas.

“I think they are great cars. I am not anti-Tesla. But I will tell you … I just really hate bad tech. And if you’ve got some bad tech that is really dangerous, I’m going to call you out on it. The problem is, do not drive your Teslas on Autopilot without paying full and absolute attention and keeping your hands on the wheel.”

The dangers:

“This phantom braking issue, where the car sees something and then decides to dramatically decelerate, that is not just a Tesla problem. We see it in many other kinds of autonomous vehicles. They’re just not reliable enough to ‘see’ the world in the way we do. We’ve done some testing with Teslas in my own lab where we can see a statistical correlation with the sun going behind clouds. Even that is enough to potentially trigger a problem with the vision system. That’s just one of the many problems, and that’s the tech problem.”

On safety features vs. convenience features:

“In automation, there are two different kinds of cars. There’s auto-emergency braking, the front collision warning, these kinds of safety devices. They are working and we can see that decreases [crashes]. But [GM’s] Super Cruise, [Ford’s] BlueCruise, these are convenience features that do latitudinal and longitudinal control for you. They’re doing acceleration for you and steering. The jury is very much out. Having come from NHTSA, I did the analysis myself on all the crash data we have, and I will tell you that if you are in an accident in a car with these convenience features, you are statistically more likely to be seriously injured or killed.”

On industry guardrails:

“There are so many good things to love about Tesla… but Tesla had some questionable design decisions about letting people be hands-free. But now all the other car companies are modeling after Tesla, and I do not think we should allow that. No car, not Tesla, not Ford, not GM, no car with any driver assist should allow you to be hands-free. Tesla’s a great car except for this bad Autopilot. When you have your hands free, it basically promotes you into complacency. So I can like the car, but not the feature.”

How long until the technology is ready for prime time:

“We’re not even close. We will see in the short term, small-mile delivery is probably where we’ll see that happen. But if you’re asking me, should I go ahead and start investing in self-driving cars because they’re going to start turning a profit next year, I don’t know when that year is going to be.”

Missy Cummings can be reached at cummings@gmu.edu.

For more information, contact Damian Cristodero at dcristod@gmu.edu or 703-993-9118.

About George Mason
George Mason University is Virginia’s largest public research university. Located near Washington, D.C., Mason enrolls nearly 40,000 students from 130 countries and all 50 states. Mason has grown rapidly over the past half-century and is recognized for its innovation and entrepreneurship, remarkable diversity and commitment to accessibility. Learn more at www.gmu.edu.

Topics

Tip Sheet

mechanical engineering

computer science

Electrical and Computer Engineering

College of Engineering and Computing

Research

Mason Autonomy and Robotics Center

Will China surpass the US in AI technology? Mason experts weigh in.

Melanie Balog — Tue, 23 Apr 2019 21:33:01 +0000

Will China surpass the US in AI technology? Mason experts weigh in. Melanie Balog Tue, 04/23/2019 - 17:33

The United States and China recognize artificial intelligence (AI) as critical to the future of the global economy and military. But the United States needs to speed up its development efforts and invest significantly more in research to remain the world leader, said George Mason University computer science professors Amarda Shehu and Craig Yu.

China released a policy note in 2017 predicting it would be the world leader in AI by 2030. A new report shows China is about a decade ahead of schedule, based on its output of highly cited research papers, which will soon be greater than those coming out of the United States.

One reason China is outpacing its prediction is because of the strategic plan of China’s State Council, Shehu and Yu agreed.

“Washington [D.C.] needs to take a page out of Beijing’s plan and accompany their strategic plan with strategic investments,” Shehu said.

China’s plan mentions more specifics on how investments will be made in AI research and technologies than America’s Artificial Intelligence Initiative does, Shehu said.

China’s investments, including their latest $30 billion venture capital fund, impact state-owned firms, academic and research institutions, commercial companies and startups, Yu said.

“With abundant funding support, government policy support and a big market to serve, some of these entities are able to grow very fast in a very short amount of time,” said Shehu, who gave the example of Sensetime, a Beijing-based company focusing on computer vision technologies that started in 2014 and is now valued at more than $4.5 billion.

But the news that China is expected to exceed the United States in AI research publications doesn’t mean they’re automatically going to be the next world leader in AI, Yu said.

“One needs to consider the impact and innovative side of research publications,” he said. “A significant amount of the most innovative breakthroughs in AI algorithms are still originating in U.S. research institutes—these U.S. institutes also have the largest pool of leading AI researchers.”

There is another challenge for the United States, Shehu said.

“Fundamental AI research happens at U.S. institutions, but it is becoming harder for these institutions to retain their best talent,” Shehu said, adding that a significant amount of AI professionals and PhD graduates choose industry jobs with lucrative benefits instead of staying in academia.

The trend could have a long-term impact as “we may not have enough professors to nurture the next generation of computer scientists and AI visionaries,” Shehu said.

“Universities need to make competitive offers to retain research talent, and it’s imperative to motivate undergraduate students to pursue graduate studies in computer science related fields, including AI,” Yu said. “This is key to maintaining a sustainable talent and innovation pipeline that propels the whole industry and research community.”

Amarda Shehu can be reached at 703-993-4135 or ashehu@gmu.edu.

Craig Yu can be reached at 703-993-4813 or craigyu@gmu.edu.

For more information, contact Mariam Aburdeineh at 703-993-9518 or maburdei@gmu.edu.

About George Mason

George Mason University is Virginia’s largest public research university. Located near Washington, D.C., Mason enrolls 37,000 students from 130 countries and all 50 states. Mason has grown rapidly over the past half-century and is recognized for its innovation and entrepreneurship, remarkable diversity and commitment to accessibility. For more information visit gmu.edu.