Re: Algorithmic TOEs vs Nonalgorithmic TOEs

From: <juergen.domain.name.hidden>
Date: Fri, 9 Feb 2001 18:34:45 +0100

Dear John,

the message was about fundamental limits of describability, not the
current limits. Descriptions are finite symbol sequences. They describe
possibly infinite things such as universe histories. Most infinite things,
however, are not describable in principle, not now, not in 2000 years.
Nobody will ever be able to fully describe anything that is not computable
in the limit by a general Turing Machine.

Algorithmic describability assumptions are stated clearly. They are
restrictive - that's why their consequences are nontrivial. The point is:
unlike your exemplary assumptions from earlier times, they are the least
restrictive assumptions we can make before we enter the nondescribable
realm where we must shut up.

Maybe algorithmic TOEs are insufficient. Once they turn out to be
inconsistent with the data we will abandon them. Unfortunately there
won't be a describable substitute.


> Resent-Date: Fri, 9 Feb 2001 07:22:51 -0800
> From: "jamikes" <jamikes.domain.name.hidden>
> Dear Juergen,
>
> thank you for your 'patient' reply. It is brilliantly worded. Let me try to
> "reword" it in a trivial and malicious way to support my point - not so
> trivial and malicious, I hope.
>
> "well, even those who don't know everything can look at all possible
> alternatives of what that everything might be"
> Possible? in their present portfolio. A trivial example: nuclear fission
> would not be among possible alternatives in 1875. Yet it was there...
> sounds like religion. God(s) are telling and priests are interpreting. Crowd
> follows.
>
> "Once they realize that the describable alternatives are restricted, they
> can derive something nontrivial from the restrictions."
> Meaning: find out something that fits your purpose and use it as "truth".
> For a Zeuss believer it was not trivial that the lightning is His anger.
>
> "The algorithmic assumption is that we don't have to consider nondescribable
> alternatives"
> Describable points to our concurrent capabilities to describe.
> We model the part we already know about and call it 'everything'. Like the
> flat earth. Organic chemistry. A-tom, not dividable further. Then make it an
> enforced science.
>
> "Maybe this assumption is wrong, but then
> we could not say anything reasonable anyway, by definition"
> A good definition: if you don't know what to say the least you can do is
> shut up.
>
> "So we shouldn't give it up unless evidence forces us"
> Play around by all means, just don't you call it science. TOE? my foot.
>
> Now the above is something completely out 0f my style. I apologize for it,
> but it was too much for me that assumptions are taken for something you can
> build on and then forget that it had an assumption for a base. It is
> ignorant, as long as you cannot prove that it is impossible to learn it (not
> in our present capabilites, however). Even in that case one should not
> assume with certainty, just confess to ignorance.
>
> Information IMO is 'difference', not imagined difference.
>
> John Mikes
>
Received on Fri Feb 09 2001 - 09:41:53 PST