I cover economics and technology for my newsletter, www.fullstackeconomics.com. Previously at the Washington Post, Vox, and Ars Technica. Subscribe to my newsletter!
Instead, we live in a world where California has Silicon Valley and anyone who claims non-competes are necessary for a healthy business climate looks a bit ridiculous.
If California hadn't banned non-competes in 1872, we could have a situation where every state enforced non-competes, everyone assumed that they must be necessary for a healthy business climate, and no state dared experiment with non-enforcement.
It's hard to overstate what good luck it was that California happened to ban non-competes more than a century ago and the courts have vigorously enforced the ban. It has provided an existence proof that an economy can work fine without non-competes.
As with the human brain you can say stuff like "well this part of the network seems to be more active when ChatGPT is thinking about X." We can modify parts of a network and observe how the output changes. But it's not a machine where you can explain its overall principles of operation in any simple or clear way.
I think the success of deep neural networks makes it pretty clear that we're unlikely to ever understand how the human brain works in any detail. Human beings built deep neural networks but there's nobody on earth who can point to a randomly-chosen cluster of neurons inside ChatGPT and explain exactly what function it serves and how it relates to other neurons.