[ExI] Bender's Octopus (re: LLMs like ChatGPT)

[Giovanni Santostasi]

*The Chinese Room argument is garbage because a magic book with the answers
to every question isn't real, and if it was, it would already be a mind.  *Yep,
basically the description of a chinese room is exactly what our brain is,
with the neurons taking the place of the people in the room. By the time
the Chinese room can answer as a sentient being then room is a mind. Not
sure why this argument was ever taken seriously.

On Sat, Mar 25, 2023 at 6:25 AM Will Steinberg via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> The Chinese Room argument is garbage because a magic book with the answers
> to every question isn't real, and if it was, it would already be a mind.
>
> I find that often thought experiments with shoddy bounds fail hard.   The
> bound here is the beginning of the experiment, where the situation is
> already magically in front of us.   Where did the book come from? How was
> it created?
>
> Of course it's easy to write out the words for a thought experiment when
> you invent an object, central to the experiment but of course not the
> subject of it, that magically does exactly what you need it to do in order
> to make the experiment.  A thought experiment could still have this book in
> it but it should be the center of the experiment
>
> On Fri, Mar 24, 2023, 5:49 AM Jason Resch via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
>>
>>
>> On Fri, Mar 24, 2023, 12:14 AM Gordon Swobe <gordon.swobe at gmail.com>
>> wrote:
>>
>>>
>>>
>>> On Thu, Mar 23, 2023 at 9:37 PM Jason Resch via extropy-chat <
>>> extropy-chat at lists.extropy.org> wrote:
>>>
>>>>
>>>> There's no way to read this paper: https://arxiv.org/pdf/2303.12712.pdf
>>>> and come away with the impression that GPT-4 has no idea what it is talking
>>>> about.
>>>>
>>>
>>> Hmm, nothing in the abstract even remotely suggests to me that GPT-4
>>> will know word meanings any more than does GPT-3. Eventually AI on digital
>>> computers will far surpass human intelligence, but even then these
>>> computers will be manipulating the forms of words and not their meanings.
>>>
>>
>> It seems to me that you have accepted Searle's arguments. I believe I can
>> knock down his Chinese room argument. If that is what you are using to base
>> your decision on you should know almost no philosophers or computer
>> scientists believe his argument holds water. Here's just one of the many
>> flaws in the argument: there's more than one mind in the room. Ask the room
>> about its favorite food, or about its experiences as a child. The answers
>> given will not be Searle's. Change Searle for someone else, the room will
>> respond the same way. Searle is an interchangeable cog in the machine. Yet
>> Searle wants us to believe only his opinion matters. In truth, his position
>> is no different than the "laws of physics" which "mindlessly" computes our
>> evolving brain state "without any understanding" of what goes on in our
>> heads. Searle's Chinese room argument works as any great magic trick does:
>> through misdirection. Ignore the claims made by the man in the room who is
>> shouting and waving his arms. Since we've established there are two minds
>> in the room, we can replace Searle with a mindless demon and there still
>> will be one mind left.
>>
>>
>>
>>> Do you believe, like my friend who fell in love with a chatbot, that a
>>> software application can have genuine feelings of love for you?
>>>
>>
>> I think we should defer such a debate until such time we can confidently
>> define what a "genuine feeling" is and how to implement one.
>>
>> Jason
>>
>>> _______________________________________________
>> extropy-chat mailing list
>> extropy-chat at lists.extropy.org
>> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230325/4d63c83a/attachment.htm>