Carnegie Mellon University is at the heart of some breakthrough technologies — from shaping the first artificial intelligence program to trailblazing the study and creation of the autonomous vehicles we see zipping around Pittsburgh’s streets today.
It has also revolutionized technology for war.
So when uniformed military officials joined CMU administrators in Lawrenceville earlier this month, kicking off a program meant to help universities and private industry to research military applications for artificial intelligence, some students and faculty weren’t exactly pleased.
Those students and faculty at Carnegie Mellon say that they are disillusioned and concerned after the school announced its partnership with the U.S. Army to develop AI solutions for the military.
Earlier this month, the Artificial Intelligence Task Force was introduced at the National Robotics Engineering Center. It’s meant as a hub for universities and private-industry partners to conduct research on AI in military applications.
While those on campus recognize CMU’s storied history with the U.S. Department of Defense — including contracting with the Defense Advanced Research Projects Agency (DARPA) on a regular basis and the hundreds of millions of defense dollars flowing into the university’s Software Engineering Institute — critics say they wish they had more information on this new work with the Army.
“We’re concerned that [the university] didn’t ask for any campus input or announce it,” said Wilson Ekern, a sophomore studying technical writing and German. “There’s a pretty big effort to get engineering and computer science students plugged into this military industrial complex.”
His sentiments come at a time when Silicon Valley and the tech industry, at large, are toeing a gray line between creating useful innovations for defense and civilian protection and producing autonomous weapons with the potential to kill.
In June, Google employees rallied against a Pentagon contract that would use artificial intelligence to interpret videos or even improve the accuracy of drone strikes, for example. About 4,000 employees signed a petition calling on the company to draft policy promising Google and its contractors would never build military technology.
The employees also alluded to the Mountain View, Calif.-based company’s former motto, “Don’t Be Evil,” in the petition, while naming other tech firms that have close ties to the military.
“By entering into this contract, Google will join the ranks of companies like Palantir, Raytheon, and General Dynamics. The argument that other firms, like Microsoft and Amazon, are also participating doesn’t make this any less risky for Google,” the authors wrote.
Since then, more than 3,100 people and 244 organizations have signed a petition pledging not to create lethal autonomous weapons. The Future of Life Institute, a Cambridge, Mass.-based nonprofit looking to mitigate existential risks created by artificial intelligence, hosts the document on its website.
Some locals have signed that petition, including Zoltan N Oltvai, associate professor of pathology at the University of Pittsburgh, and Spenser Neel, a student there.
Still, Department of Defense contracts are lucrative.
In 2017, it spent more money on federal contracts — $320 billion in current dollars — than all other government agencies combined, according to The Congressional Research Service, a Washington, D.C.-based public policy research arm of the U.S. Congress.
A significant portion of those defense dollars are spent on research and development at universities.
“More than half of DOD’s basic research budget is spent at universities and represents the major contribution of funds in some areas of science and technology,” per the report.
Before the AI task force was even a thought, technology had already been changing the face of war. Leading-edge technologies have been developed to save soldiers’ lives, for example.
Drones make it possible for forces on the ground to remotely control autonomous air vehicles, with no pilot on board. Smart bombs — controlled by radio or guided by lasers, with a built-in computer — allow targets to be hit with precision, reducing casualties.
Gen. John Murray, commander of the Army Futures Command, which is geared toward modernizing the military, said at the press conference earlier this month that in the near-term, he imagines facial recognition technology could aid in combat.
Still, when he applied to the university, Mr. Ekern of CMU said that “they don’t tell you they’re also a great war school.”
Mr. Ekern is also recruitment chair for Students for a Democratic Society, a student group “fighting against the militarization of research,” according to its Facebook page.
In September, that group protested at an on-campus conference where military contractors like Lockheed Martin and General Dynamics Mission Systems were recruiting students, reports the university’s student newspaper, The Tartan. About 10 members passed out fliers that stated “some of the attending institutions are, simply put, bad for humanity.”
The student group also recently published an op-ed in The Tartan, decrying the Army AI Task Force and a lack of transparency.
“Even our very own university president, Farnam Jahanian, is committed to war over diplomacy and negotiations. ... This is unacceptable. The president did not seek the input of students, faculty, or staff before proffering this view on our behalf — or indeed, before committing Carnegie Mellon to the Army AI Task Force itself,” the group wrote in its opinion.
“For our part, SDS members stand ready to hold the U.S. military accountable and check its power at our university,” it continues.
“When President Jahanian states that AI military applications ‘benefit society,’ we strongly disagree. Any task force designed to keep the U.S. at war forever, dominating and controlling the world, is anathema to democracy, peace, and social justice.”
Carnegie Mellon’s direct and indirect ties to the military are troubling to Mr. Ekern.
“You hear little whispers here and there about things that come out of the university,” he said. “You don’t really understand the level of how much the university works hand in hand with the military ... to make it easier to kill people.”
During the press conference earlier this month to debut the Army AI task force on campus, CMU President Jahanian noted that the university has a long history of sponsored research with the Department of Defense, but that it was unclassified work in the public domain, with both defense and civilian applications.
Robert Strauss, professor of economics and public policy at CMU, wrote in an email that the AI task force represents a shift in the university’s academic culture.
“When I arrived [in 1979], I learned that classified research for [the Department of Defense] was not part of the intellectual culture or practice of the university, and there were, at that time, vocal voices among the faculty, and voices representing the faculty ... who spoke out against it,” he said.
Michael McQuade, vice president for research at CMU said in a statement that he expects early work will be non-classified.
“It is expected that the initial projects and technologies created within the Task Force will be non-classified, and the findings will be available for study and use by the scientific community and the public at large,” he said in a statement. “The Task Force's immediate focus is on predictive maintenance and automated logistics.”
He went on to say that faculty and students have “flexibility” to choose whether or not they want to participate in any of the task force’s research.
To be sure, CMU is not the only university with ties to the military.
The Massachusetts Institute of Technology has had its Lincoln Laboratory on campus since 1951. It’s federally funded and run by MIT on behalf of the Department of Defense. Mr. Strauss points out that lab has done research on nuclear weapons.
When Mr. Jahanian was pressed by reporters earlier this month about whether the university had set up an ethics committee to evaluate a partnership with the U.S. Army, he did not directly answer, but instead offered that the task force would create a forum for discussions on ethics of AI in a military context.
That concerns Mr. Strauss.
“That the new tie ... occurred without internal CMU discussion [that I am aware of] shows that the line has blurred now,” he said. “It reflects a continued, accelerating decline in the faculty governance in the university.”
As for Mr. Ekern, he feels concerned that students’ opinions aren’t part of the conversation.
“It seems like they were doing everything they could to avoid feedback from students or faculty,” he said. “Why do our opinions count less than the administration in accepting these deals?”
Courtney Linder: clinder@post-gazette.com or 412-263-1707. Twitter: @LinderPG.
First Published: February 18, 2019, 12:30 p.m.
