> Are patents (and trade secrets) the right model for protection?

This probably deserves a more expansive discussion in its own right. I think what has become clear is that the tools we have (patents, trade secrets, copyright) are extraordinarily crude. Nonetheless, they worked pretty well in a time when intrinsic friction and overhead in markets was high.

Now that we’ve eliminated almost all friction in markets, everything has become extremely sensitive to the intrinsic cost and supply chain structure of the IP being protected and the speed at which those supply chains can adapt to competitive opportunities. To make matters worse, these intrinsic cost and supply chain structures tend to vary considerably depending on the particulars of the domain. Consequently, the IP protection tool in any particular case is the one that in practice has highest cost burden (both money and speed) on anyone trying to circumvent it with the lowest cost to both implement and defend. Patents rarely win that argument anymore for many things.

The most robust protection these days is information asymmetry, and trade secrets are good at maximizing that. Unlike patents, it is also extremely cheap to implement, so it is economically favorable. Defending that information asymmetry has become more difficult in some ways (more sophisticated reverse engineering) but easier in others (hello cloud). Unfortunately, we lived in an age of trade secrets in past centuries and it had many downsides; patents were essentially invented to mitigate those downsides.

The entire dynamic has many downsides but trade secrets are essentially winning out. Information asymmetry still has a lot of market power if you are good at exploiting it, much of the benefit isn’t even the secrecy per se in many cases.