> On 10 Oct 2018, at 07:50, Brent Meeker <meeke...@verizon.net> wrote: > > > > On 10/9/2018 9:45 PM, Philip Thrift wrote: >> >> >> On Tuesday, October 9, 2018 at 8:16:59 PM UTC-5, John Clark wrote: >> On Tue, Oct 9, 2018 at 7:54 PM Pierz <pie...@gmail.com <javascript:>> wrote: >> >> >I refuse to accept that "axiom", and I also do not feel compelled to >> >embrace solipsism. >> >> You are able to function is the world so you must have some method of >> deciding when something is conscious and when it is not, if its not >> intelligent behavior what is it? >> >> > I think it is entirely possible - and indeed sensible - to believe that >> > some entities that behave "intelligently", like the chess app on my >> > iPhone, are insentient. >> >> I don't know what the quotation marks in the above means but if something >> acts intelligently then it is sensible to say it has some degree of >> sentience. >> >> > Whereas some entities that behave unintelligently (like Donald Trump >> > (sorry, I really shouldn't)) are sentient. >> >> I admit it's a imperfect tool but it's all we've got and all we'll ever have >> so we just have to make good with what we have. A failure to act >> intelligently does not necessarily mean its non-sentient, perhaps both a >> rock and Donald Trump are really brilliant but are just pretending to be >> stupid. If so then both are conscious and both are very good actors. >> >> > The absence of an objective test for third-party sentience does not force >> > one into solipsism. It may point to 1) a problem with your ontology >> > (qualia aren't "real") >> >> That means nothing. I detect qualia from direct experience and that outranks >> everything, it even outranks the scientific method; so if qualia isn't real >> then nothing is real which would be equivalent to everything being real >> which is equivalent to "real" having no meaning because meaning needs >> contrast. >> >> > or 2) a deficient state of knowledge wth respect to the (pre) conditions >> > of consciousness. >> >> I don't know what that means either. >> >> > Seeing as you have no theory of consciousness at all, >> >> Yes I do. My theory is that consciousness is the way data feels when it is >> being processed and that is a brute fact, meaning it terminates a chain of >> "why is that?" questions. >> >> > statements like "you have no alternative but to..." don't have much force. >> > There are plenty of alternatives, >> >> Name one! I ask once more, in you everyday life when you're not being >> philosophical you must have some method of determining when something is >> conscious, if its not intelligent behavior what on earth is it? >> >> > a refusal to engage it as a problem, in spite of the increasingly >> > widespread acceptance among scientists that it is a real problem, and >> > possibly the biggest problem of all in our current state of knowledge >> >> I think intelligence implies consciousness but consciousness does not >> necessarily imply intelligence, so the problem I want answered is abut how >> intelligence works not consciousness. >> >> John K Clark >> >> >> >> One could look at it that way. In terms of biological evolution, what has >> turned out to be intelligent beings (us!) are also conscious beings. When we >> started making computers and programming languages and such (inventing a >> field called Artificial Intelligence), it got a little confusing. Is IBM >> Watson [ https://en.wikipedia.org/wiki/Watson_(computer) >> <https://en.wikipedia.org/wiki/Watson_(computer)> ] "intelligent"? Some >> might say yes, others, no. There are some AI scientists (or SI - Synthetic >> Intelligence, to contrast with AI [ >> https://en.wikipedia.org/wiki/Synthetic_intelligence >> <https://en.wikipedia.org/wiki/Synthetic_intelligence> ] who say to make >> truly intelligent artifacts they must be conscious. >> >> So the question remains no matter how one parses intelligence and >> consciousness: How do you make a conscious robot? > > I'm obviously not sure, but here's an idea of how consciousness might occur > based on Jeff Hawkins ideas in his book “On Intelligence”. I refer to the > intuition pump of an AI Mars Rover: > > <fdnhpkhninkblnhk.png> > > The sensors of the MR would define the current status, both internal and > external. This goes into a predictor that estimates how the current status > will change if there's no change in the current plan. The prediction from > the previous cycle is compared to the new current status. If there's not > significant difference, it's “Ho Hum” and action proceeds as planned. But if > the comparison shows a deviation from expectation That is something to take > note of. It's noted in long-term memory which is a searchable database which > can be used to learn from. And it initiates a need to update the plan. So > what rises to the level of consciousness is something that is surprising and > may need a change of plan. And if you ask the MR what happened, it will > refer to it's long term memory to give an account based on what it saw as > significant > > Brent > P.S. That rainbow pyramid thing is a hierarchy of values per Maslow that are > used in evaluating a plan. > > This comports with the idea that conscious thought is a kind of post-hoc > commentary on what you're thinking, and explanation you can tell yourself and > other people. Remember that one of the criticisms of neural nets is that they > don't explain themselves.
Re-entrant multiple layer are sort of universal learner. We have still to train them to try to explain themselves. As far as they remain self-referentially in that process, they will converge on the theology of the self-referentially correct machine, described by G*, and close the the “neoplatonist”. They will understand that the hard problem of “matter” is harder than the hard problem of consciousness, as it requires more complex quantitative prediction. > That means if you want an explanation from an ANN it has to be a separate > function which can also be implemented by some more NN. But then you have no > guarantee that the explanation is the real one. Right, but that is the case for all explanations, not just the metaphysical one. That is why scientists avoid terms like “true”, real”. With mechanism we could argue that even “infinite/finite” should be avoid to, except at the meta-level. And similarly in metaphysics we can talk of “truth" and “real” as object of research, and object of clarification in the frame of some theory/assumption. But there too, the notion of true or real cannot be invoked. Some religious people sum up the idea by “God has no Name”, or “the Tao which has a name is not the Tao”, or “what we cannot talk about we should not talk about”, but of course, this is already saying too much. Yet Lao-ze get it right when he said “the wise stay mute, the fool talk”. It is the difference in the truth, by definition, for a correct machine, of ~[]<>t (which belong to G* minus G), and <>t -> ~[]<>t, which belongs to G, and the neural nets training on re-entrance, with enough layer, cannot avoid the difference between many ways to access some truth, well before even inventing the language to be able to justified a little part of it. Biological evolution "has favoured” neural nets, but the neurons are quite sophisticated bio-molecular technology and biochemical machines, and the receptor are subject to genetical regulation, so the neural nets itself evolves from a myriad of genetical algorithm. Rover need sex, make a family, …, to evolve. it needs taxes and death for a long period before getting the correct idea where its consciousness comes from, and if matter is an illusion or not, without ever being sure of anything in that domain, as long as attached to its “material” appearance. Bruno We are not human beings having spiritual experiences, we are spiritual beings having human experiences. (de Chardin). > > Brent > > -- > You received this message because you are subscribed to the Google Groups > "Everything List" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to everything-list+unsubscr...@googlegroups.com > <mailto:everything-list+unsubscr...@googlegroups.com>. > To post to this group, send email to everything-list@googlegroups.com > <mailto:everything-list@googlegroups.com>. > Visit this group at https://groups.google.com/group/everything-list > <https://groups.google.com/group/everything-list>. > For more options, visit https://groups.google.com/d/optout > <https://groups.google.com/d/optout>. -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To post to this group, send email to everything-list@googlegroups.com. Visit this group at https://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/d/optout.
