Re: The philosophy of computation reformulates existing ideas on a new basis --- EQUIVOCATION

Am Sat, 02 Nov 2024 20:33:40 -0500 schrieb olcott:

On 11/2/2024 8:22 PM, Richard Damon wrote:

On 11/2/24 9:00 PM, olcott wrote:

On 11/2/2024 7:52 PM, Richard Damon wrote:

On 11/2/24 8:38 PM, olcott wrote:

On 11/2/2024 7:21 PM, Richard Damon wrote:

On 11/2/24 5:13 PM, olcott wrote:

On 11/2/2024 3:24 PM, Richard Damon wrote:

On 11/2/24 12:56 PM, olcott wrote:

On 11/2/2024 10:44 AM, Richard Damon wrote:

On 11/2/24 8:24 AM, olcott wrote:

Of course, that is for this exact input, which uses the copy of H
that does abort and return.
>

No it is not.
   when HHH simulates DDD(), it's analyzing an "idealized"
   version of DDD() where nothing stops the recursion.

In other words you are admitting that it isn't actually looking at
the input it was given.

ChatGPT (using its own words) and I both agree that HHH is supposed
to predict the behavior of the infinite emulation on the basis of
its finite emulation.

LLMs literally string words they have previously seen together.

Yes, but that behavior is DEFINED by the actual behavior of the
actual machine.

No it is not. It is never based on the actual behavior of the actual
machine for any non-terminating inputs.

Haha what? It absolutely is. For a nonterminating input a halting
decider must return that it doesn't halt.

Then you don't undetstand the requirement for something to be a
semantic property.

The actual behavior specified by the finite string input to HHH does
include HHH emulating itself emulating DDD such that this DD *not some
other DDD somewhere else*

Especially not some DDD that calls a non-aborting simulator HHH1.

cannot possibly reach its own "return" instruction whether HHH emulates
DDD forever or some finite number of times.

DDD returns, and you need to give the encoding of this DDD to HHH.

I can see this, Ben can see this and ChatGPT understands it so well
that it can use entirely different words to explain exactly how it
sees this.

Nope, If you look carefully at what Ben agreed to was if you define the
NON-SEMANTIC property that you have been trying to define, your decider
can be a correct POOP decider. (of course, you can't look that closely
as you don't undetstand what you have been talking about).

The semantic property of this finite string does specify that HHH must
emulate itself emulating DDD.
The direct execution of DDD DOES NOT SPECIFY THAT.

Yes of course it does. It calls HHH.

--
Am Sat, 20 Jul 2024 12:35:31 +0000 schrieb WM in sci.math:
It is not guaranteed that n+1 exists for every n.