[ExI] My guesses about GPTs consciousness

Jason Resch jasonresch at gmail.com
Sun Apr 16 19:51:27 UTC 2023


On Sun, Apr 16, 2023, 3:23 PM Brent Allsop via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

>
> Yes Rafal,
> Most everything you say seems quite insightful, and helpful.
> I'd also like to know: would you define consciousness?
>
> On Sun, Apr 16, 2023 at 6:59 AM SR Ballard via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
>> On Sun, Apr 16, 2023 at 1:25 AM Rafal Smigrodzki via extropy-chat <
>> extropy-chat at lists.extropy.org> wrote:
>>
>>> 4) I do not know what structures arose in the GPT4 to be able to process
>>> sensory (e.g. visual) information. If they are different from the human
>>> sensory cortex, the corresponding qualia might be also completely different
>>> from human.
>>>
>>
> This sounds very compatible with the way I think about qualities of
> subjective knowledge.
> Here is the image I always show to describe my view:
> [image: 3_robots_tiny.png]
> In other words, robots represent knowledge of red things with abstract
> names, like the word 'red'.  The system can't know what redness is like,
> and it can't know what the word red means, without a dictionary.
> For us, the subjective quality of our knowledge of red things is just a
> fact.  No dictionary required.
> So, just wondering if you would agree with this?
> And the supporters of the Representational Qualia Theory
> <https://canonizer.com/topic/88-Theories-of-Consciousness/6-Representational-Qualia>
> petition camp are working on building and tracking consensus around a
> definition of consciousness as:
> *"Computationally bound elemental subjective qualities like redness and
> greenness."*
>
> Oh, and Jason had a great idea of how to improve this image, so here is an
> early draft (we're still working on making it better):
> [image: The Strawberry is Red_02.jpg]
> Jason, let me know what you think. Does it better communicate what I'm
> trying to say about qualities of conscious knowledge?
>
>
Yes! This is a massive improvement!

I especially like that it reveals the internal perceptions within their
heads. It might benefit  to label the three people so that you can
distinguish them more easily in a textual description. Like A, B, C, D.

Or alternatively, to give the three people distinguishing features (E.g.
make one a man, or have a different hair style and give a hat to one of
them, or something). That way you can refer to cases more easily.

Jason



>
>
>
>
>
> _______________________________________________
> 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/20230416/15ea2ce7/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 3_robots_tiny.png
Type: image/png
Size: 26214 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230416/15ea2ce7/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: The Strawberry is Red_02.jpg
Type: image/jpeg
Size: 34123 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230416/15ea2ce7/attachment-0001.jpg>


More information about the extropy-chat mailing list