How do we compensate for centuries of human wisdom?

You can hear the doubt in their voices when they say they don’t owe anything to the humans who wrote the words they trained the machines on. You hear the towing the party line. You hear the company march, left, right, left. They don’t buy it, but they hope we do. 

They say the machines learn the same way as us: we read lots of things, our knowledge base grows, and then we regurgitate from the pages we’ve read, sculpting it into something slightly different. It’s just your average process of learning, they say. Nothing new here. 

They trained the AI on everything the New York Times has ever published and then quickly tiptoed away, hoping it would go unseen. Hoping the billions of dollars the Times has invested over the past 173 years to build a trove of human history could be theirs for the snatching. Hoping they could philosophize their way to a free lunch. 

Ezra asks what will happen when the AI has put all the journalists out of a job and there is no human to feed the AI its in-the-world information anymore. How will it know what’s happening on our plane, so it can explain it back to us, if there is no one acting as its eyes and ears to begin with? Everyone just shrugs.