.... An AI system that can do the sorts of things Jim does toIt would be no use if the system could prove that it knows math, but could not
prove his claim that he knows math would be quite useful IMO.
test others for their knowledge of math. Knowledge representation of some form
would be required here, because it would have to take into account the differences in the ability to use math, set at appropriate limits beyond which
no testing is done. For example, it wouldn't test Jim on whether he can multiply 6 digit numbers in under a fraction of a second, and it would have to
test him for abilities that are peculiar to human beings, but only such as pertain to their use of computers, not such as they evince among themselves. Perhaps 'knowledge representation' is not the best term to use, since the abilities it tests in human beings or in anything else are defined by humans themselves, so in effect it would be more of an identification process rather than its "knowing" something. Likewise, it could not really test a human for its knowledge of math, but could for example "know" that Jim is working on a certain math formula and keep track of his progress. Now it doesn't seem possible that if friend of Jim is working on a similar formula that it could investigate areas of application for the formulas to work together. The symbols
used in algebraic formulas are specific to the formulas, and the purpose of the
formula is to represent a change in that which the symbols represent. Change is inherent in the nature of things and computers calculate the nature of the change, but changes cause things to transform themselves into each other so that nothing remains of their previous state, therefore how can a computer use
certain formulas to derive new ones that bear no resemblance to them?
--
http://www.costarricense.cr/pagina/ernobe
On Thursday, December 30, 2004 at 3:56:13 PM UTC-8, ernobe wrote:
.... An AI system that can do the sorts of things Jim does toIt would be no use if the system could prove that it knows math, but could not
prove his claim that he knows math would be quite useful IMO.
test others for their knowledge of math. Knowledge representation of some form
would be required here, because it would have to take into account the differences in the ability to use math, set at appropriate limits beyond which
no testing is done. For example, it wouldn't test Jim on whether he can multiply 6 digit numbers in under a fraction of a second, and it would have to
test him for abilities that are peculiar to human beings, but only such as pertain to their use of computers, not such as they evince among themselves.
Perhaps 'knowledge representation' is not the best term to use, since the abilities it tests in human beings or in anything else are defined by humans
themselves, so in effect it would be more of an identification process rather
than its "knowing" something. Likewise, it could not really test a human for
its knowledge of math, but could for example "know" that Jim is working on a
certain math formula and keep track of his progress. Now it doesn't seem possible that if friend of Jim is working on a similar formula that it could
investigate areas of application for the formulas to work together. The symbols
used in algebraic formulas are specific to the formulas, and the purpose of the
formula is to represent a change in that which the symbols represent. Change
is inherent in the nature of things and computers calculate the nature of the
change, but changes cause things to transform themselves into each other so that nothing remains of their previous state, therefore how can a computer use
certain formulas to derive new ones that bear no resemblance to them?
--2022 Update: I guess you should look for your epistemology elsewhere.
http://www.costarricense.cr/pagina/ernobe
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 297 |
Nodes: | 16 (0 / 16) |
Uptime: | 128:50:21 |
Calls: | 6,663 |
Calls today: | 1 |
Files: | 12,212 |
Messages: | 5,335,293 |