On Tue, Apr 15, 2014 at 11:42 PM, meekerdb <[email protected]> wrote:
> On 4/15/2014 1:41 PM, Telmo Menezes wrote: > > > > > On Tue, Apr 15, 2014 at 6:44 PM, meekerdb <[email protected]> wrote: > >> On 4/15/2014 4:38 AM, Telmo Menezes wrote: >> >> An interesting related hypothesis is that language originated >>> from synesthesia caused by psychadelics. >>> >>> Telmo. >>> >>> >>> I had heard that Telmo. Do you have a reference, a link? >>> >> >> Unfortunately not. I think I heard in a talk. Might be related to >> McKenna's "stoned ape" theory, but I can't find anything... >> >> >> That seems very far-fetched considering that animals already exhibit >> rudimentary language and that its selective advantage for a tool making >> social animal is huge. >> > > I agree that the idea that language was bootstrapped by psychadelics is > far-fetched. I see it as a fun hypothesis more than anything else, for the > reasons you mention. > > >> I don't see how synesthesia could do anything but confound and confuse >> the development of language. >> > > Maybe so for the development of direct symbols, but I can imagine it > playing a role in the emergence of more abstract ideas. Even in modern > times we can see this at work, to a degree. Many of the cultural ideas that > originated in the 60s, and that still reverberate today, were "unearthed" > by using LSD, cannabis, etc. > > > What cultural ideas would those be? Get out of Viet Nam? Civil rights > for blacks? The pill? > > > > I find the effects of psychoactive substances particularly interesting > for AI research, because they show a profound way in which our brains > differ from the current model of computation. Computer programs typically > crash if we mess with their computational substrate. We flood the brain > with an inhibitor for a certain type of receptor or with the analogue of > some transmitter and it doesn't collapse. It does all kinds of interesting > things, some good and some bad. Sometimes you get "the dark side of the > moon" -- if musical talent is already present, of course :) > > > I think the analogy is wrong. Brains compute by chemical transmitters. > So when we interfere with the chemistry, its analogous to changing program > steps in a digital computer - not to messing with the substrate (e.g. > silicon). > Here I meant the substrate as the von Neumann model -- which is reflected in modern computer languages. In any case, messing with a transistor, a memory bit, the compiler or the source code mostly results in the same kind of critical failures and almost never leads to a different or interesting mode of operation. > A brain is a neural network. It can (probably) be simulated by a digital > computer; but the simulation will be a low level. At that level LSD would > be simulated as changing some connection strengths. > It can be argued that a computer program is a function network. It could also be argued that a neural network is also a function network. I would say that these things are incidental, and the big deal is the network topology, and the algorithms that lead to the topology. Artificial neural networks or not, we don't really know how to produce functional networks with the same type of adaptability that we observe in the brain, nor do we really know how to do general-purpose computing outside of the von Neumann model (or maybe lambda calculus, with the old lisp machines). Even changing things a bit, like what happens with modern GPUs, we lose generality. Telmo. > > Brent > > > Telmo. > > > -- > 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 [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/everything-list. > For more options, visit 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 [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/d/optout.
