3 ms·

The naive explanation is that it helps inventors, because speculative investors (the trolls!) buy up patents. One analogy is pharma research. Rights for a prom

by pas 2y ago

The naive explanation is that it helps inventors, because speculative investors (the trolls!) buy up patents. One analogy is pharma research. Rights for a promising candidate molecule are purchased by "big pharma" and they will do the grunt work to validate it and extract the big money from its therapeutic value. Substitute "FDA market authorization after successful clinical trials" with "that infamous East Texas court district and picking the right targets" (picking a too big target might backfire, picking a too small doesn't really worth the costs, etc.) Of course the questions are: does this really help inventors? do inventors need help? is it good for society that inventors get help? is the cost of helping inventors this way not unreasonable to the economy? and even if the cost is "low", how fucking fair is it that a lot of businesses are using a given invention but only a lucky few get dragged to court? can we do better? what kind of people patent trolls are? what do they do with the money?

returningfory2 2y ago
Yeah, I think pharma is the really tricky case. I worked at a small biotech company whose business model was (in part) to do early stage drug discovery research and then sell promising leads to bigger companies, who would take the leads through the FDA approval process. Actually taking a drug through to approval is a $1 billion+ endeavor (with a high probability that it just won't work; e.g. stage 3 trials just fail). Small companies cannot do this. So a naive solution like "don't allow patents to be sold" actually restricts a bunch of reasonable businesses.
- ajb 2y ago
  There's an actionable difference between pharma and most other things. Property rights should "give notice" of the boundaries of a property, most obviously you lose certain rights if you fail to mark the edge of real property with a fence or something. But for most patents there's no realistic way to find patents you might be infringing (for software this is impossible in the general case as a corollary of Fines theorem). That's why we hate patents so much: you only find out if there is a patent if you get sued. But for pharma there is a way; chemicals are "indexable": because a standard mapping from chemical structure to names exists, if you are using a chemical and want to find whether anyone patented it, you can do so easily. In an ideal world, the law would recognise this and patents would only be valid in areas where there is a natural way of indexing - which is pretty much only pharma AFAICT

by pas 2y ago

returningfory2 2y ago

Yeah, I think pharma is the really tricky case. I worked at a small biotech company whose business model was (in part) to do early stage drug discovery research and then sell promising leads to bigger companies, who would take the leads through the FDA approval process. Actually taking a drug through to approval is a $1 billion+ endeavor (with a high probability that it just won't work; e.g. stage 3 trials just fail). Small companies cannot do this. So a naive solution like "don't allow patents to be sold" actually restricts a bunch of reasonable businesses.

ajb 2y ago
There's an actionable difference between pharma and most other things. Property rights should "give notice" of the boundaries of a property, most obviously you lose certain rights if you fail to mark the edge of real property with a fence or something. But for most patents there's no realistic way to find patents you might be infringing (for software this is impossible in the general case as a corollary of Fines theorem). That's why we hate patents so much: you only find out if there is a patent if you get sued. But for pharma there is a way; chemicals are "indexable": because a standard mapping from chemical structure to names exists, if you are using a chemical and want to find whether anyone patented it, you can do so easily. In an ideal world, the law would recognise this and patents would only be valid in areas where there is a natural way of indexing - which is pretty much only pharma AFAICT