AGI

Just ran across a field c asked artificial general intelligence, lilac very interesting. Some good points made about computability, and simplifying assumptions. Curious to me that recognising the imposition that context inevitably makes, no one seems particularly interested inseeing that as fundamental.

Published with Blogger-droid v1.7.1

Running grumbles

I did something terrible to my left shin in the 16K race last weekend, in which I took a minute off my PB, which, frankly, is rubbish. I was planning to knock of a good 8 minutes, but I'm blaming a nasty twinge in my hip flexor (right) for the failure. Irritatingly though, I first got the twinge at training on a day where I thought I'd got a slightly longer stride working really well. ([500, 2:15, 150] as opposed to [500, 2:05, 170]).

Anyway, 9 days, NINE, of rest is what it took to get back on the track. And I was 8% slower, and 8% more out of condition when I did. And, the flexor is still twinge-y. Grrrr. But at least the shin splints pulled up OK.

Axiom 97

Language predisposes, it does not essentialise.

(Gee, I wonder what I was thinking about when this popped up?)

Perhaps it was this: when we (any two-plus people) talk about something, we approach it through series of approximations; we negotiate ourselves to a shared* understanding, which may still be incomplete because there probably isn't time to fully check all the separate ramifications, each of which has it's own ramifications, etc. etc. So we are hoping that our (by definition incomplete) mutual understanding is sufficient and relevant.

That thing named is not made into an essence through the naming. It is not, through being named, made fundamental, made into something requiring consent from all future speakers. They remain free to negotiate, although, generally, the results of past negotiations require, as they age and attract support, increasing energy and good fortune to succeed in making a change.

Take democracy. There was a time when that meant men voting for MPs. Then the "men" part turned out not to be important. No epistemically interesting word escapes this problem.

If we take "existence" to mean "existence for people" (try and remove every trace of "people" from a statement in a human language) then essence is at best "having been named". For example, a "tree" is a "tree-as-conceived-by-people". It's not a "tree-as-conceived-by-borer".

*Not at all this simple

more jogging thoughts

- exercising the machinery of creativity does not manufacture truth. (imagine a couple of film-ic types, ya know, we need something shocking here, how about the guy amputates his own leg with a saw, nah, nah, how about he amputates his father's leg with a whacking great knife & lets him bleed to death, yeah cool, right, then he fucks his mother... and based on this in the year 4000 they construct a theory of human behaviour - reckon it'll be any good?)

- Deep things are hard to understand, and the struggle to understand them can provoke strong emotions; but it does not follow that things which are hard to understand are necessarily deep (comment on the legions of inept disciples of Foucault et. al.)

- Rhythm is patterns of cause and effect transformed into patterns of time (I was thinking about the way in which you can do a lot of things once you have organised a pattern for them, so you don't need to think about the nature of the interrelationships all the time. The interrelationships are still all there, but the ordering of the rhythm of your life incorporates them automatically)

- Thinking costs 20 seconds per half-kilometre. Time to run .5K thinking about stride, breathing & arms is 2:20. Thinking about work/life/study the time rises to 2:40. Mind/body, anyone?

Fwd: Musicality of prose (1)

Acquisition, retention and use : rhythm ../. ./. . /       ; assonance frustrated

Retention, acquisition and use :rhythm ./. ../. ./        ; assonance frustrated

Retention, acquisition, utilisation :rhythm ./. ../. .../.  ; assonance fulfilled

Not quite a filler

This is really a note to self - not sure if I've made this point before, but i want to lodge an objection to the subject, or at least the idea that all sentences necessarily have a subject (as per the LFG crowd), or that subjects are necessarily agents (Baker) (assuming causation is at stake), or that the subject is a "grammaticised topic" (also LFG). my intuition is that these are all very eurocentric, possibly latincentric.

I'll follow that up when I get back to a library, but I haven't seen remotely testable evidence for any of this. I also find the failure of anyone to produce a principled account of the operation of the subject in Chinese - which however clumsily has nonetheless been plausibly challenged by Lee.

You know...

I just want to warn any readers - not that I have undue expectations - that my expertise in these matters is not necessarily great. Relatively great perhaps. Greater than many. But not great. it's hard to get great into a modest sentence, but I assure you that's what I was aiming at.

Knowledge is tricky. In fact, it's a minefield. Unbelievably, given that Plato draws our attention to Socrates' attempts at consciousness-raising in this regard a good two point five plus millennia ago, I have to feel that the philosophy department hasn't been doing particularly well on this question. Of course, there is another way to look at it; the philosophy department is still being funded to sort this question out, so they have been successful in fending off the need to pursue any new questions to think about for quite a while. If your idea of a stressful day is a nice quiet think, then obviously you don't need to be bothered with the stress of working out what to think about it.

It seems to me that the best overview of the problem is that we know something when we have sufficient evidence for it; that seems perfectly clear until you have to define sufficient and evidence & I'm certainly not going to try. My own feeling is that we only ever know things inside a closed framework & that the foundations of that framework, the axioms, are inevitably matters of faith. That's how it is. It's irritating for the scientists, but it shouldn't be. It makes the god-botherers feel superior to the scientists (that's what annoys the scientists really, no-one like being looked down on), but it shouldn't. There are many things that could be used as a basis for comparison between the two frameworks, (scientific & god-bothered) and in fact I imagine there's a sufficient plethora to support on-going dissent for a few more millennia. It's wouldn't hurt if we could raise the standard of debate though - eh, Richard, now, would it?

One thing that puzzles me though, and that's how it is that my pseudo-scientific colleagues in the linguistics faculty imagine that "falsification" is the last word in scientific methodology. Frankly, I haven't even found much evidence that they can do that successfully - there are honourable exceptions, mostly in the, guess where, cognitive science faculty - but even if they could, falsification isn't such a big deal. Popper is usually credited with it as a response to the problem of empiricism, which, simply put, is that induction requires the future to resemble the past so that the assumptions made from past experiment can be held as future laws. Falsification purports to do away with this by positing a one-time event, falsification, which mysteriously underwrites the future. Of course falsification relies on the future being the same as the present just as much as induction. Plus, as Noam demonstrates admirably through most of the sixties, a bit of countervailing evidence is no obstacle to a good theory. There are a large number of political and rhetorical strategies to ignore it. One thing is for sure; if linguistics IS a science, then it provides some of the best evidence that science is another human activity dependent largely on the vagaries of human behaviour; a love of fashion, a desire to be popular, the pursuit of power.

I like linguistics & it has achieved a lot; predicting population movements well in advance of DNA testing, for example. I like science & it would be insane to trivialise its achievements or its methodology. "Science is bunk", he typed into his computer wirelessly connected to the internet. Ho, ho, ho.

I don't like all the claims made for knowledge: it's too stressful. I haven't got the time to unpick the contingency in everything I study - but the authors have. That would be a kind of scholarship that I would like to see.