Have researchers abandoned work on AI that thinks like a human?
Can we put our hopes (or fears) of a computer that thinks like a human to rest?
Perhaps, according to Andrew Moore, dean of Carnegie Mellon University's School of Computer Science.
Moore, in an interview with Axios, said artificial intelligence researchers have essentially stopped working on creating a machine that thinks like we do and instead are focusing on what they've learned about AI thus far.
“We have pretty much stopped trying to mirror human thinking out of the box. We are focusing on engineering (what) has already been invented,” Moore told Axios.
So what has been invented? A lot. Moore mentioned AI's ability to interpret speech and play difficult games. CMU is home to some of the top speech recognition and machine translation researchers in the world, and a poker-playing AI developed at the university has beaten some of the top players.
And while AI researchers might be taking a step back from the goal of artificial general intelligence — a machine that thinks like a human — working on what's already been discovered “will lead to big advances,” Moore said.
Moore and CMU President Farnam Jahanian traveled to the White House last month to take part in a summit on artificial intelligence with members of the president's staff and executives from Google, Amazon, Intel, Ford and more. The university announced in May that it will be the first in the country to offer an undergraduate degree in artificial intelligence.
CMU is considered the birthplace of artificial intelligence and has the top AI graduate program in the country, according to rankings by U.S. News and World Report. Most recently, two researchers signed on to work at Facebook's new AI lab in Pittsburgh part time and JPMorgan hired Manuela Veloso, the former head of CMU's Machine Learning Department, to head AI research at the bank.
Aaron Aupperlee is a Tribune-Review staff writer. Reach him at firstname.lastname@example.org, 412-336-8448 or via Twitter @tinynotebook.