The AI Panic is Insane

Both arrows and cars can get out of hand, and make us go extinct

By Deirdre N. McCloskey.

"Maybe not in Brazil, but in the USA we are going insane about artificial intelligence. Our Congress might act to regulate, which is always a worrying sign of a panic attack.

Artificial intelligence is the ability of a computer with root to do things commonly associated with intelligent beings, such as us. The science fiction is that AI will take us over. Hundreds of technology leaders signed a letter recently declaring that "Mitigating the risk of extinction from AI should be a global priority alongside other societal scale risks such as pandemics and nuclear war."

Such panic is insane.

Doing things commonly associated with intelligent beings, after all, is what human culture has done from top to bottom since the beginning. The philosopher Alfred North Whitehead remarked in 1911 that " Civilization advances by extending the number of important operations which we can perform without thinking of them." Correct.

The bow and arrow, for example, is a "computer with robot" for throwing a little spear called an arrow without thinking much. It takes over part of the aiming that a human would commonly do throwing a spear. The earliest punched-card computers in Belgium for making rugs were computer-controlled robots making it unnecessary for human to think about what color of weft to cast through the warp. All physical tools do jobs commonly associated with humans. The mechanisms for, say, driving are such computers.

You might argue, correctly, that both arrows and cars, though presumably not rugs, can get out of hand, and make us go extinct. But that of course is true for many novelties—say, the atomic bombs that the terrified tech leaders were referring to. Or war ships using wind to make thoughtful rowing unnecessary, and bloody naval war easier.

And think beyond physical tools. Brazilian music, cuisine, economy evolve unpredictably, for better or worse. The Portuguese language is a non-physical tool, of human manufacturer though not of human design. The old humanist joke is, "Do we speak the language... or does the language speak us?"

Language is in fact probably the most dangerous artificial intelligence machine ever. For example, technology leaders can use scary language to drive us insane.

Bring in state regulation of language, eh? No."