Re: Comp and causality
On 29 Nov 2012, at 17:07, Roger Clough wrote: Hi Bruno Marchal I think of comp as a monitor of what the brain does physically, objectively, materialistically, such as, at first glance, produce electrical signals, things a computer implant could do and in fact do do. But the brain also operates biochemically, not just electrically, so I suppose one would need a biochemical computer. Could that be done ? We don't need that. Nothing in biochemistry is known not to be Turing emulable. Is the computer causally connected to the brain. ? The brain is the (natural) computer, like the heart (the biological organ) is a ntural pump. I don't see how this could be possible. It would have to know when a woman's period is, and its hormonal changes, the effects of age, etc. ? It would be so much more plausable if comp only monitored, not controlled, the brain's activities. Causality is much much more difficult. That could come later if at all. "Causality" is a high level concept, usually captured by some modal logic (and they are an infinity of them) in the form of the necessity of an implication [](p->q). Bruno [Roger Clough], [rclo...@verizon.net] 11/29/2012 "Forever is a long time, especially near the end." -Woody Allen - Receiving the following content - From: Bruno Marchal Receiver: everything-list Time: 2012-11-22, 09:27:20 Subject: Re: Reality Check: You Are Not a Computer Simulation [Audio] Hi Roger, On 22 Nov 2012, at 11:25, Roger Clough wrote: Hi Bruno Marchal You say " OK, but invalid when used to pretend that we are not machine, like Penrose and Lucas did." So basically, whether you believe the Lucas-Penrose theory It is not a theory. It is an informal argument according to which Gödel's theorem would show that we are not machine. The argument has never convinced any logicians and can be shown wrong in many different ways. On the contrary, incompleteness protects the consistency of Church thesis, and thus comp. depends on whether you believe in comp or no. Not at all. The argument show that Gödel's theorem (incompleteness) ==> non-comp. This would imply that comp ===> Gödel's theorem is wrong, which is absurd. The most basic error is that Lucas/Penrose believe that a human can know that they are sound. Like Watson can play jeopardy, Gödel already knew that the Löbian machine can detect the error made in Penrose and Lucas type of argument. This is developed in my long text: "Conscience & Mécanisme". Judson Webb wrote a book on this. In his second book, Penrose correct his mistake, but does not really take the correction into account, and thus miss the formal first person indeterminacy. I have serious problems with comp because the 1ps and hence the 3ps of various people and various computer programs will vary. I don't see how they can all be the same. I don't understand your point. Meanwhile, I'll look at the counter-arguments to Lucas and Penrose. You need to study Gödel's incompleteness theorem. Most popular account of it are non valid. An nice exception is Hofstadter "Gödel, Escher Bach". [Roger Clough], [rclo...@verizon.net] 11/22/2012 "Forever is a long time, especially near the end." -Woody Allen - Receiving the following content - From: Bruno Marchal Receiver: everything-list Time: 2012-11-21, 12:23:40 Subject: Re: Reality Check: You Are Not a Computer Simulation [Audio] On 21 Nov 2012, at 11:32, Roger Clough wrote: Hi Bruno Marchal I'm trying to understand your paper, but a seemingly much simpler form of your argument keeps getting in the way. The simpler form is the Lucas argument, discussed in great scholarly detail on http://www.iep.utm.edu/lp-argue/ To be franc there is nothing new in that paper, on the contrary it fails to mention the work done by Webb (not to talk on mine on Lucas, Benacerraf and the Penrose argument). I have counted more than 50 errors in Lucas paper. Some are uninteresting, and some are very interesting. The argument of Lucas and Penrose are typically invalid, but it can be corrected, and it leads to the proposition according to whioch: If I am a machine, then I cannot know which machine I am, and this plays some role in the formal part of the study of the first person indeterminacy. Lucas and Penrose assumes that they are sound, and that they know that they are sound, but this is already inconsistent, even if the soundness is restricted to arithmetic. In Conscience & Mechanism, I show how all Löbian machines can refute Lucas and Penrose. Basically they confuse []p (3p beliefs) with []p & p (1p knowledge).. It seems to me to be self-evident that 1p cannot be part of 3p But that is good insight of you. For correct machine, this can be proved, as the machine cannot prove the true equivalence between []p and []p & p, as they don't know that they are correct. []p can be defined in the language
Re: Comp and causality
On Thu, Nov 29, 2012 at 5:07 PM, Roger Clough wrote: > Hi Bruno Marchal > > I think of comp as a monitor of what the brain does > physically, objectively, materialistically, such as, at first > glance, produce electrical signals, things a computer > implant could do and in fact do do. > > But the brain also operates biochemically, not > just electrically, so I suppose one would need a biochemical > computer. Could that be done ? > The computation medium is not relevant. Any algorithm that can be executed by the biochemical computer can also be represented in a digital computer. In fact a lot of people are endeavouring to do just that, namely in the field of artificial neural networks. The trouble is figuring out what the algorithm is. > > > Is the computer causally connected to the brain. I don't see how this > could be possible. It would have to know when a woman's > period is, and its hormonal changes, the effects of age, > etc. > Why the dualism? Why not see those things as part of the computation? > > It would be so much more plausable if comp only monitored, > not controlled, the brain's activities. Causality is much much > more difficult. That could come later if at all. > > > [Roger Clough], [rclo...@verizon.net] > 11/29/2012 > "Forever is a long time, especially near the end." -Woody Allen > > > - Receiving the following content - > *From:* Bruno Marchal > *Receiver:* everything-list > *Time:* 2012-11-22, 09:27:20 > *Subject:* Re: Reality Check: You Are Not a Computer Simulation [Audio] > > Hi Roger, > > On 22 Nov 2012, at 11:25, Roger Clough wrote: > > Hi Bruno Marchal > > You say > > " OK, but invalid when used to pretend that we are not machine, like > Penrose and Lucas did." > > So basically, whether you believe the Lucas-Penrose theory > > > It is not a theory. > > It is an informal argument according to which Gödel's theorem would show > that we are not machine. The argument has never convinced any logicians and > can be shown wrong in many different ways. > > On the contrary, incompleteness protects the consistency of Church thesis, > and thus comp. > > > > > depends > on whether you believe in comp or no. > > > Not at all. The argument show that Gödel's theorem (incompleteness) ==> > non-comp. This would imply that comp ===> Gödel's theorem is wrong, which > is absurd. > > The most basic error is that Lucas/Penrose believe that a human can know > that they are sound. > > Like Watson can play jeopardy, Gödel already knew that the Löbian machine > can detect the error made in Penrose and Lucas type of argument. This is > developed in my long text: "Conscience & Mécanisme". Judson Webb wrote a > book on this. > > In his second book, Penrose correct his mistake, but does not really take > the correction into account, and thus miss the formal first person > indeterminacy. > > > > I have serious problems > with comp because the 1ps and hence the 3ps of various > people and various computer programs will vary. > > I don't > see how they can all be the same. > > > I don't understand your point. > > > > Meanwhile, I'll look at the counter-arguments to Lucas and Penrose. > > > > You need to study Gödel's incompleteness theorem. Most popular account of > it are non valid. An nice exception is Hofstadter "Gödel, Escher Bach". > > > > > [Roger Clough], [rclo...@verizon.net] > 11/22/2012 > "Forever is a long time, especially near the end." -Woody Allen > > > - Receiving the following content - > *From:* Bruno Marchal > *Receiver:* everything-list > *Time:* 2012-11-21, 12:23:40 > *Subject:* Re: Reality Check: You Are Not a Computer Simulation [Audio] > > > On 21 Nov 2012, at 11:32, Roger Clough wrote: > > Hi Bruno Marchal > > I'm trying to understand your paper, but a seemingly much simpler > form of your argument keeps getting in the way. The > simpler form is the Lucas argument, discussed in great > scholarly detail on > > http://www.iep.utm.edu/lp-argue/ > > > To be franc there is nothing new in that paper, on the contrary it fails > to mention the work done by Webb (not to talk on mine on Lucas, Benacerraf > and the Penrose argument). > > I have counted more than 50 errors in Lucas paper. Some are uninteresting, > and some are very interesting. The argument of Lucas and Penrose are > typically invalid, but it can be corrected, and it leads to the proposition > according to whioch: > > If I am a machine, then I cannot know which machine I am, and this plays > some role in the formal part of the study of the first person indeterminacy. > > Lucas and Penrose assumes that they are sound, and that they know that > they are sound, but this is already inconsistent, even if the soundness is > restricted to arithmetic. > > In Conscience & Mechanism, I show how all Löbian machines can refute Lucas > and Penrose. Basically they confuse []p (3p beliefs) with []p & p (1p > knowledge).. > > > > > It seems to me to be self-evident that > > 1p cannot be part of 3p > > > >
