[ExI] Peer review reviewed AND Detecting ChatGPT texts

Adrian Tymes atymes at gmail.com
Tue Jan 10 00:58:59 UTC 2023

On Mon, Jan 9, 2023 at 9:06 AM Brent Allsop via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> If a robot is going to follow logic, no matter how much you teach it to
> claim it has qualia, you can prove to it it can't know what redness is like.

As noted below, I think we may be using "robot" to mean different things
here.  A chat bot that can not see, has no direct knowledge of redness.
Chat bots are not the only kind of robot in existence.  Indeed, despite
having the word "bot" in their names, chat bots are not the first kind of
thing that comes to my mind upon hearing the word "robot".

> Perhaps you haven't seen my "I Convinced GPT-3 it Isn’t Conscious" writeup
> <https://docs.google.com/document/d/17x1F0wbcFkdmGVYn3JG9gC20m-vFU71WrWPsgB2hLnY/edit> where
> I've done this 3 times.

I read it over.  The logs don't appear to show what you say they show.
Even the title is not correct: "I do not have consciousness in the same way
that humans do" (which you referenced in your email, but I am directly
quoting from the log for emphasis) is not the same as "I am not
conscious".  At most, you got one of the chatbots to admit that it is only
a language model, and as such does not have vision or vision-related
experiences, such as perceiving red.  That is that specific instance, not
all robots in general.

> Oh, and could we release chatbots into Russia, to convince
> Russian citizens how bad the war in Ukraine is?  Or could Putin do the same
> to the US and European citizens?

There is reason to believe that Russia has been doing this for some years
now.  I would hesitate to credit Putin directly, but some people in his
administration may well have done so.

> I bet logical chat systems will always support the truly better side more
> successfully than anyone trying to get people to be bad, or non
> democratic.  In other words, I bet an intelligent chat bot that would help
> Kim Jong-un would be far more difficult than building a chat bot that
> supports bottom up democracy, and so on.

I suspect you are quite right on this, at least when those chat systems are
talking to rational people.  There is a problem when those systems are
designed to press on things that make people irrational.  The last few
lines from the chat bot in the second transcript in your paper starts to
show what I mean: nonserious talking that appears to be intended more for
emotional impact than logical.  Another example, mentioned elsewhere in
this email thread, is the potential for religious evangelist chat bots.

Part of the problem is that evil is so simple to support as opposed to
good.  A farmer may fear that not giving most of his crops today means his
family will be made into tomorrow's publicly tortured then executed
example.  This requires very little thought, as opposed to planning how to
convince all the farmers in his village not to give in and thus keep the
would-be thugs hungry enough that they are unable to brutalize anyone.  But
note the influence of fear - not an element of rationality - in the
farmer's thinking.

Granted, a mere chat bot that controls no physical body is no more able to
wield a rifle than to experience redness.

> On Sun, Jan 8, 2023 at 9:44 PM Adrian Tymes via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>> On Sun, Jan 8, 2023 at 8:24 PM Brent Allsop via extropy-chat <
>> extropy-chat at lists.extropy.org> wrote:
>>> The AI can't know what redness is like, while we do  know, infallibly,
>>> what redness is like.
>> Can't it?  By the argument you have often cited, it can't know our
>> version of redness, but it would know its own version of redness.
> Chat bots represent red things with an abstract word like 'red' which, by
> design, isn't like anything  Good logical chat bots can be forced to
> agree with this.
> <https://docs.google.com/document/d/17x1F0wbcFkdmGVYn3JG9gC20m-vFU71WrWPsgB2hLnY/edit>
> Are you claiming that the word 'red' somehow results in an experience of
> redness?

I believe the disconnect here is what sort of AI is meant.  I refer to AI
in general.  I agree that a chat bot with no visual system can not know red
like we can.  However, I believe that a similar AI that included a visual
system with color sensors, like the robot depicted in your red apple/green
apple/machine apple image, could have its own version of redness.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230109/c7ad8af8/attachment.htm>

More information about the extropy-chat mailing list