Speaker cautions the future of artificial intelligence

Documentarian James Barrat lectured on artificial intelligence in the Modlin Center for the Arts in front of a crowd of roughly 200 people yesterday as part of the Weinstein-Rosenthal Forum on Faith, Ethics and Global Society. Barrat bleakly depicted how the unintended consequences of artificial intelligence, or AI, could potentially devastate the human race.

AI refers to computer systems that perform tasks usually requiring human intelligence. In recent years, large organizations such as Google, Apple and the National Security Agency have poured billions of dollars into researching and developing an AI equivalent to that of a human being, Barrat said.

“Virtual brains at the price of a computer will be the most lucrative invention in modern history,” Barrat said, but he warned that they were a serious gamble for mankind because of their destructive power.

AI that are equal to the mind of a person, which many experts agree will be possible around 2030, would have the same creative abilities as a human being. These AI would then build other types of AI, creating a “super AI.” Super AI would be higher functioning than the human brain and smarter than people. Much of Barrat’s lecture revolved around the dangers of having superintelligent AI.

Many experts, such as renowned physicist Stephen Hawking, find this possibility extremely dangerous because of the AI’s ability to design sophisticated weaponry. Super AI would strictly adhere to logic unless current developers begin researching how to design moral values into computers. With super AI only acting logically, they would begin a process of self-preservation with no regard for compassion or kindness.

The next 20 or 30 years are critical to solving this issue before it becomes a problem, Barrat said.

“For all intents and purposes, they will be 100 percent amoral, super intelligent sociopaths,” Barrat said. He said he would rather see research go toward studying the human brain first before developing AI that would be difficult to control.

“The people with the deepest pockets aren’t interested in governing it at all,” Barrat said. “We have an opportunity now but our window is closing. We have to figure out what it is about ourselves that we want to exist in the future.”

The moral values that should be encoded into AI are not easily identified because they vary depending on culture, location and timeframe, Barrat said.

Barrat said it was difficult not to draw on Hollywood when discussing the possibility of machines overtaking the human race, but stressed caution in using movies as evidence, where mankind always wins in the end. “We may not have the solutions,” Barrat said.

The window of this issue is certainly closing fast. IBM’s supercomputer, Watson, has completed tasks that baffled its creators, Barrat said — a sign of how advanced AI is becoming. Watson has already won on the TV show “Jeopardy” and is studying to become a digital medical assistant.

“My ethics class focuses on AI in the military, so it was nice to hear about how they are focusing on AI in a civilian setting,” said Jennie Trejo, a freshman who was excited that Watson was working in medicine.

At its core, Barrat’s topic asks the question of whether human beings have intangible qualities that cannot be attained through technology, or whether the brain is simply another biological tool replicable with science. While Barrat said he believed science would eventually match the computing power of humans, not everyone in attendance agreed.

“I think that people are more than the sum of their bits and bytes. There’s an aspect of our spirit and existence that’s not present in machines,” said Elizabeth Dwyer, Westhampton College ’76. “AI can be creative, but we are creative in a fundamentally different way.”

Those in attendance certainly absorbed Barrat’s hesitations about artificial intelligence.

“I thought machines would become better and better but not necessarily move beyond human intelligence,” said Bill Burch, a Richmond community member. “As he talked about it, it became more and more of a real possibility that they will.”

Barrat convinced many in the crowd that Super AI is on the horizon.

“It’s probably inevitable,” said Carole Weinstein, who invited Barrat to speak at Richmond and is one of the main sponsors of the forum. “Hopefully, it’s miraculous.”