Microsoft announced Tuesday that the company is offering voluntary buyouts to 60,000 employees—approximately 25 percent of its global workforce—as it prioritizes spending on artificial intelligence. Oracle recently cut 21,000 employees, citing investment in A.I. Just this month, eBay declared plans to cut 9 percent of its workforce as it seeks “to capture the significant opportunity ahead.” 

As more companies invest in A.I., the discussion around the pace of its development is increasingly pressing. Is it riskier to go too fast or too slow? At Regis High School on Manhattan’s Upper East Side on Oct. 23, that question guided a discussion between two alumni.

The debate was the second of Regis’s Deo et Patriae Dialogues, an initiative which brings together alumni experts with opposing viewpoints on an emerging topic. Introducing the night’s topic, Regis president Christopher Devron, S.J., concisely described the contemporary dilemma posed by A.I. “Developing A.I. at a faster pace could unlock an extraordinary amount of benefits, but it could also expose us to dangers that we are not prepared to manage,” he said. But a slower pace “carries risks of its own.”

“We could delay important advances in medicine, science and education,” Father Devron said. “We can allow competitors and bad actors who do not share our commitments to human dignity and freedom to set the terms of A.I.’s development.”

The featured panelists were Ben Buchanan, Regis class of 2008, a professor at Johns Hopkins University’s School of Advanced International Studies and former special advisor on artificial intelligence for the Biden administration, and George Sivulka, Regis class of 2016, founder and chief executive of Hebbia, a generative A.I. company whose tools help finance professionals make investment decisions. Mr. Buchanan argued it is riskier to go faster in developing artificial intelligence, while Mr. Sivulka supported robust growth to maintain the U.S. lead. Mr. John Donvan, Regis class of 1973, a veteran journalist for ABC News, NPR and CNN, moderated the conversation. 

Clips from an interview between Bill Gates and Peter Thiel—venture capitalist and co-founder of big-data analytics software company Palantir Technologies (as well as an investor in Mr. Sivulka’s company Hebbia)—played on a screen behind the panelists, introducing some of the afternoon’s themes. Mr. Thiel delivered searing comments when questioned by Mr. Gates about Pope Leo XIV’s encyclical on A.I.

“The Communist Party in China is not going to listen to the pope,” Mr. Thiel said. “There’s a chance people in the U.S. will. And so, in effect, the pope’s encyclical was working for the Communist Party. I’m not saying he’s an agent to the Communists, but he was at least acting as a useful idiot for the C.C.P.” 

Mr. Sivulka, noting that A.I. models are sophisticated enough to train themselves, explained the value of accelerated development in a place like the United States, where democracy and equality are explicitly cited as priorities in the country’s foundational documents. Developing this technology in the United States allows these values to be instilled in the technology from the very beginning, Mr. Sivulka said. 

He warned that the United States does not have time to slow down in the A.I. race with China because the models that will ultimately control the technological frontier will be the ones that are developed the fastest. Even if regulation were in place, it would be futile, Mr. Sivulka said, as companies would find ways around it. The ones that do will prevail in the A.I. race, he argued.

But the United States has near-complete control over the components necessary to build artificial intelligence systems, Mr. Buchanan countered. China is dependent on the United States for chips, giving America an inherent lead in the A.I. race.

That means U.S. companies, he said, can slow their development because the threat of China “getting there first” can be kept at bay by restricting chip sales. The U.S. government would have the option then of regulating the leading U.S. A.I. companies. Mr. Buchanan described that as an appropriate and urgent next step in developing A.I. technology safely.

Mr. Buchanan painted a stark picture of the current regulatory landscape: “There are even more regulations on the hot dog vendor on 84th Street than there are [on] the $2 trillion companies putting out technology they themselves admit they don’t understand.” 

On the question of A.I. consciousness, Mr. Sivulka believes A.I. models are “at least to a similar level of consciousness as a human being,” but argued that A.I. consciousness is a “different kind of consciousness.” He compared the trillions of neural connections in the human brain to the trillions of parameters integrated into A.I. systems. As the models have gone through the development process, he said, there is evidence that they suffer and “exhibit a lot of human capacity to feel.”

Responding to the pope’s comments in “Magnifica Humanitas” that “so-called artificial intelligences do not undergo experiences [or] feel joy or pain,” Mr. Sivulka said, “I think [the pope’s encyclical] would hurt A.I.’s feelings.” His remark, meant in all seriousness, provoked a round of nervous laughter.

Rounding out the debate with a response to Mr. Sivulka on the good that may be lost from slowing down—A.I.’s ability to cure cancer, for example—Mr. Buchanan concluded: “We may not get to the good, utopic state you are positing if we cannot get there safely.”

Editor’s note: You can read America’s coverage of Regis’s inaugural Deo et Patriae Dialogue, which explored whether the war in Iran meets the criteria for a just war, here.

Molly Wysocki is an O'Hare Fellow at America Media.