I was profoundly struck by the Wall Street Journal’s reporting from Amrith Ramkumar, Angel Au-Yeung and Lindsay Ellis on the week AI anxiety went mainstream. Without downplaying legitimate concerns about the havoc that new technologies such as AI can cause, history offers some interesting analogies.  

Historically, new technologies that dramatically increased the speed, scale, or autonomy of human capability have triggered intense anxieties. Disruptive technologies such as AI, in the Scott Anthony sense, make what was once difficult or that required expertise or skill easy. As a consequence, they make those same things cheap. Combine those two phenomena and the rules of life change utterly. Unsurprisingly, those likely to be negatively affected protest (the Luddites never did regain their former standard of living). Equally unsurprisingly, the rest of us wonder what it means for us. Below, I unpack some of the key issues and look at historical analogies for perspective.

Those concerned about AI as a new species, a kind of superintelligent entity, suggest that it could quite unintentionally (from its designer’s perspective) develop objectives fundamentally misaligned with human survival, rendering us powerless to stop it.  Some call this the “alignment problem.”

In 1858, a New York Times article (rediscovered by this piece in The Atlantic from 2014 ) argued that the telegraph was making information transmission “too fast” for the truth. Specifically, “Superficial, sudden, unsifted, too fast for the truth, must be all telegraphic intelligence. Does it not render the popular mind too fast for the truth?

Source link