Peggy Noonan complains that our country never decided to "do" AI ("History Raises Its Eyebrows at the Week in AI," Declarations, Oct. 17). There was no vote, no debate. This is true, but neither were there votes on automobiles, nuclear energy and weapons, electricity, computers, the internet or social media.
Had our government not continued developing nuclear weapons we would probably be using a Cyrillic alphabet today. To avoid the catastrophe Ms. Noonan expects, we need the brightest and most capable people devoted to intercepting and neutralizing any villainous cyber weapons our many adversaries are now using against us and will clearly choose to use with more potent AI weaponry. Artificial Intelligence dangers won't stop simply because we put the brakes on.
Larry W. White
Dallas
---
I am a coder and am working with some of the best AI models available today. Not one of the models "thinks" like any engineer I work with. They are getting better at simulating thinking for longer and longer before veering off into lunacy. But they always require a human mind watching them and correcting their mistakes. Long term they will only ever amplify their users' intelligence, not replace it.
John Mullan
Oxford, Pa.
---
If I am properly understanding Ms. Noonan's warning, she is saying AI is at the stage comparable to that of the first full scale test of an atomic bomb at Trinity in 1945. The great fear then was that the test may prove catastrophic, setting the atmosphere on fire, but it went ahead anyway. The government deemed the risk worthwhile considering the millions of lives projected to be lost in an invasion of the Japanese mainland. If that's where we are now with AI, is the risk offset by the potential reward? What would Harry Truman do?
Richard J. Carhidi
New York
(END)