Т.В. Черниговская недавно обмолвилась о статье, где в качестве примера выступает умный пылесос:

Я развлеку вас под конец, недавно появилась замечательная статья, которая называется «Где границы моего сознания?».  Автор очень интересно пишет: «Купил я пылесос с искусственным интеллектом, который катается. Я смотрю на этот пылесос, он разумный, и всё», – я читала и страшно смеялась. «Ведёт себя просто как живой, внуки мои его печеньем кормят, он печенье это ест. Смотрю я на него и думаю: наверно, есть в нём репрезентация. Открыл», – тут я потеряла сознание, – «Открыл, нет там никаких  репрезентаций, только кошачий пух и мусор. Закрыл, и не знаю, что теперь делать».

Я так понимаю, что речь шла о рецензии «Where is my mind?» Джерри Фодора. По крайней мере, пример этот он там приводит, вкупе с еще разными интересными размышлениями. Рецензия написана на книгу Энди Кларка «Supersizing the Mind: Embodiment, Action and Cognitive Extension». Насколько я понял, основная мысль книги заключается в том, что можно считать, что наш разум воплощается в том числе во внешних объектах. Т.е. поскольку не доказана принципиальная разница между записной книжкой и нашей памятью, то почему бы не принять тот факт, что записная книжка — тоже часть разума, как и память:

A. Clark: Such considerations of parity, once we put our bioprejudices aside, reveal the outward loop as a functional part of an extended cognitive machine. Such body-and-world involving cycles are best understood . . . as quite literally extending the machinery of mind out into the world – as building extended cognitive circuits that are themselves the minimal material bases for important aspects of human thought . . . Such cycles supersize the mind.

Фодор с этим не согласен на основе двух положений:

  • при принятии этой идеи вводится представление о дополнительном процессе «вспоминания о вспоминании», излишнее для объяснения памяти как таковой; соответственно, этот процесс позволяет различить внутреннее и внешнее в случае памяти.
  • есть разница между первичной (underived) и вторичной (выведенной, derived) репрезентацией, где первые являются признаком разума, а вторые лишь строятся на основе разума

Кларк в ответном комментарии приводит аргумент «последовательного перехода»: заменяя последовательно части мозга на искусственные мы не можем определить границу, где внешний мозг переходит во внутренний. Это вкратце, подробнее смотрите в самом комментарии. Забавно, что как раз про этот аргумент Фодор пишет «Я предлагаю по доброте душевной не придираться к опасным переходам».


Понравившиеся мне цитаты:

Let me tell you about my new vacuum cleaner. Unlike my old vacuum cleaner, this one is a sort of robot. You don’t have to push it around to get the dirt up: it’s got wheels and it rushes from place to place, vacuuming wherever it happens to be. Which path it decides to take determines what it happens to pick up. It decides what path to take at a given point by executing a ‘random walk’. If it bumps into something (the couch as it might be), the shock of the bump causes the robot to turn an arbitrarily determined number of degrees in an arbitrarily selected direction. It then proceeds to vacuum some more and continues to do so, da capo, until somebody remembers to turn it off. From time to time, it gets trapped behind things, or under things, and sometimes it gets tangled up in a fringe; but, by and large, it works pretty well. The rugs do get cleaner. And my grandchildren adore it. What they particularly like is feeding it. They put bits of detritus in its path, which the robot then duly gobbles up. The cat disdains the thing but the rest of the family is rather fond of it.

Is what my robot does when it ‘decides’ to change course a sort of thing which if it had happened inside the robot, ‘I would have had no hesitation in accepting as part of [a] cognitive process?’ The trouble with this way of phrasing the issue is that, in the crucial cases, one doesn’t know how to apply it. That’s because what one thinks about the parity principle itself depends on what one thinks about EMT. If it’s your view (as I guess it’s mine) that mental events are ipso facto ‘internal’, then you will, of course, deny that something that happens on the outside could be mental. It’s irrelevant whether what the vacuum cleaner did would have counted as making a decision if the robot had done it on the inside because, according to mind/world dualists, making a decision isn’t the kind of thing that can happen on the outside. Clark would doubtless say that this begs the question against EMT, and he would be right. But a mind/ world dualist would say that the parity principle begs the question too, only in the opposite direction. And, anyhow, what on earth are we talking about? I guess I understand the thesis that my vacuum cleaner changed course because it collided with the couch in my living-room. But how am I to understand the hypothesis that it would (or wouldn’t) have changed course if it had collided with the couch in my head? All that it literally has inside is rug dust and cat hair: I know because I’ve looked. There isn’t room for a couch in my vacuum cleaner; or in my head.

… slippery-slope arguments are notoriously invalid. There is, for example, a slippery slope from being poor to being rich; it doesn’t follow that whoever is the one is therefore the other, or that to insist on the distinction is mere prejudice. Similarly, there is a slippery slope between being just a foetus and being a person; it doesn’t follow that foetuses are persons, or that to abort a foetus is to commit a homicide.

… The mark of the mental is its intensionality (with an ‘s’); that’s to say that mental states have content; they are typically about things. And (with caveats presently to be considered) only what is mental has content. It’s thus unsurprising that considerations about content are most of what drives intuitions about what’s mental. For example, Clark (and Heidegger) notwithstanding, tools – even very clever tools like iPhones – aren’t parts of minds. … What I should have said isn’t that only what’s literally and unmetaphorically mental has content, but that if something literally and unmetaphorically has content, then either it is mental (part of a mind) or the content is ‘derived’ from something that is mental. ‘Underived’ content (to borrow John Searle’s term) is the mark of the mental; underived content is what minds and only minds have. Since the content of Otto’s notebook is derived (i.e. it’s derived from Otto’s thoughts and his intentions with a ‘t’), the intensionality of its entries does not argue for its being part of Otto’s mind. So the intensionality of notebooks can be granted by someone who doesn’t think that notebooks are the sorts of thing that could be parts of minds.

… Say, if you like, that my vacuum cleaner uses the world itself as its model of the world; my vacuum cleaner does, after all, change direction when it bumps into the couch, and the couch is, after all, in the world. But then, my vacuum cleaner is very stupid.

http://www.lrb.co.uk/v31/n03/jerry-fodor/where-is-my-mind