The reliable determination of novelty is an intractable problem. Empirically, this appears to be true; anecdotally, i know of at least one patent office that h

The reliable determination of novelty is an intractable problem. Empirically, this appears to be true; anecdotally, i know of at least one patent office that has tried to reliably determine novelty and failed, and i know of none that have succeeded. To justify regulation of innovation, the burden of proof should be on those who think that determination of novelty is feasible to show via demonstration that it actually is. Why is this intractable? Technical knowledge is continually ramifying into an ever-finer tree of subspecialities. Imo a significant proportion of researchers at the cutting edge of knowledge work in areas whose details are only understood by a small number of people in the entire world (~5-30). Someone outside of this small group will not be able to reliably distinguish novel ideas from non-novel. Yet it is economically infeasible for a government to hire one patent examiner for every one of these subspecialties (this would require the patent office's budget to be a significant fraction of the sum of all research budgets in the entire world). Therefore any real-world implementation of the patent system will always have an unacceptably high number of bad patents because insufficently specialized examiners will allow patenting of things which are obvious to the (handful of) actual specialists in the given area. Pharmaceuticals appear to be a special case in that there is a more crisp delineation between chemicals than between ideas.