[ExI] e: GPT-4 on its inability to solve the symbol grounding problem

[Gordon Swobe]

To put that another way, just as it can know the forms of words but not
their meanings, it can know the form of grammar but not its meaning. I have
no doubt that this goes a long way toward helping it write sensibly. The
forms of grammar are part of what is meant by the patterns of language.



-gts


On Sun, Apr 16, 2023 at 1:03 AM Gordon Swobe <gordon.swobe at gmail.com> wrote:

> On Sat, Apr 15, 2023 at 10:47 PM Jason Resch via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
> I disagree completely.
>
> The LLM can learn patterns, yes, (this is part of what GPT does and
> reports that it does) but no, it cannot learn what the parts of speech
> mean, nor does it claim to know the parts of speech.
>
> > Numbers are often listed in a certain sequence (e.g. in ascending
> order), often used before nouns (can infer it gives a count).
>
> What is a number? What is a noun?  A person, place or thing, you say? What
> is a person, a place, or a thing?
>
> > Words that can stand alone as sentences are verbs.
>
> What is a verb? Action? What is action? For that matter, what is a
> sentence?
>
> Likewise with all other parts of speech. Even if it can classify every
> noun as belonging to a certain set of symbols X, and every verb as
> belonging to another set of symbols Y, it could still never know what is a
> noun or a verb. It can know only the pattern of how these classes of
> symbols tend to appear together .
>
> GPT-4 can learn only the patterns and relationships between and among
> word-symbols with no knowledge of the meanings of the individual words, *exactly
> as it reports that it does*. It does this extremely well, and it is in
> this way that it can *simulate* human understanding.
>
>
> -gts
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230416/712a46b7/attachment.htm>