Delirium wrote:

Ray Saintonge wrote:

...

I would call Sabine's comment realistic rather than pedantic. I realize that technophiles dream of these artificial solutions, but the best they can realistically hope for is competent interpretation of technical material. When it comes to metaphor technophiles are playing in the right field of dreams. (I wonder how a machine would translate that into the language of people who know nothing about baseball.)

Off-list since this is getting off-topic, but I was implying that my comment was pedantic---I agree with Sabine as a practical matter, but she was arguing that in principle it is *impossible* for machine translation to meet the quality of human translation, which I think is incorrect. It is, however, correct that it will not do so in the near future.

I don't know about it being off-topic. It does have a bearing on how we approach Africa. If machine translation were effective it would make the task of going into Africa a lot easier. I simply do not share your unbounded optimism about the future of technology. Of course, in the nightmare scenario there will always be the HAL who will need to be reminded that he should stick to "A Bicycle Build for Two."

The essential observation is that humans are, according to one line of thinking, themselves nothing more than biological machines. Therefore, by definition, anything a human can do is something a machine can do, because we already have an example of a machine that can do it. All that's left is the engineering problem of how to build such a machine ourselves. It could either be a workalike (e.g. wait for neuroscientists to document the exact functioning of the brain, and simulate that in a computer), or else it could be something radically different that still does things we'd consider as requiring intelligence.

The premise that humans are biological machines does not have unanimous support. What you say may very well follow logically from that premise, but others will see it as a kind of GIGO argument

The "proof" for this as in principle possible is by contradiction: If we assume it's impossible, that means that humans must be something other than a glob of matter that obeys the laws of physics. Thus, there must be something magical about humans. That's an assumption many people aren't willing to make, which means that they must then therefore accept that humans are in essence machines, albeit fantastically complex ones.

The glob of matter as matter may very well obey the laws of physics, but which physics? Does that necessarily exclude obeying additional laws? The leap to magic, like the leap to faith, involves jumping to conclusions. I'm just as unwilling as you to make that leap, but that still does not lead me to the conclusion that you derive. Rather it leads me to a lot of unanswered questions. What, for example, is the role of intuition? I'm not prepared either to accept my own speculative theories as an answer, or to confuse that speculation with objective truth. That would be engaging in pseudoscience. I am, however, prepared to live with the uncertainty that some questions may not be answerable. I think that a proof by contradiction can only work if ALL the contradictory possibilities are known and have been explored.

Ec