On Wed, Sep 11, 2019 at 6:10 PM Jose Mario Quintana <[email protected]> wrote: > Raul also wrote: > > ((Honestly, though, I am surprised that you were not already aware of > > this aspect of SAT-3. How could SAT-3 have an equivalence to SAT-n > > without multiple clauses?)) > > :D I do not recall stating anything about, or even mentioning, a k-SAT > problem, at all. Do you? I do remember quoting from the paper which > mentions only the 3-SAT problem; see below. (Do not blame me, I am just > the messenger.)
Let's try this again: Let's say we've got a relatively small expression, with only 50 variables. And, let's say we express that in terms of SAT-3, with three variables per clause. Do you honestly believe that there would only be one clause (only one order, in the market representation)? Do you have reason to believe that the clauses would be independent of each other (which would free us from regulations against interdependent orders)? If so, could you give an example of how this could work? > Is the forum's number one fan of the "Markets are efficient if and only if > P = NP" paper now trying to attack it? If so, then maybe this is an > indication that the poor dead horse has been beaten long enough. You made the claim that what you call "circuit breaker" regulations would not be relevant in the context of using the market as a SAT-3 problem solver. I feel that that claim was bogus, because a typical SAT-3 expression will involve many interdependent clauses. ... I it's pretty clear that a SAT-3 solver involves many interdependent clauses ... Do you see why I think that this regulatory issue is significant? Do I need to keep restating the point until I find a way of expressing this, which you can understand? Or do you sincerely think that SAT-3 solvers don't involve interdependent clauses? > No? Then hurry up, program the market and run it, executing suitable OCO > orders as necessary, before the restless regulators storm TD's facilities > and shut down their WebBroker platform, This suggests to me that you also did not understand the reasoning expressed in the paper. The "program the market" idea involved a large investment and the payout was getting the solution to an otherwise intractable problem. That's only a money-maker if the intractable problem was in the way of achieving something valuable. But since it's exceedingly unlikely that the market can solve these kinds of problems, there's not much likelihood of getting the payout from that investment. Worse, since there are regulations (which the author of the paper probably wasn't aware of) which would prevent the "market programming" from functioning, the likelihood of payout plummets. Now, maybe -- as you seem to have suggested -- those regulations would not be enforced. That would take us from essentially no chance of payout to an exceedingly infinitesimal chance of payout. Thanks, but no thanks, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
