Connectionists: Chomsky's apple

[shpandex at tuta.io]

On Wed, Mar 08, 2023 at 09:25:02AM -0800, Terry Sejnowski wrote:
> I have always been impressed with Chomsky's ability to use plausible
> arguments to make his case even when they were fallacious.
>
> https://www.nytimes.com/2023/03/08/opinion/noam-chomsky-chatgpt-ai.html#commentsContainer
>
> "Here???s an example. Suppose you are holding an apple in your hand. Now you let the apple go. You observe the result and say, ???The apple falls.??? That is a description. A prediction might have been the statement ???The apple will fall if I open my hand.??? Both are valuable, and both can be correct. But an explanation is something more: It includes not only descriptions and predictions but also counterfactual conjectures like ???Any such object would fall,??? plus the additional clause ???because of the force of gravity??? or ???because of the curvature of space-time??? or whatever. That is a causal explanation: ???The apple would not have fallen but for the force of gravity.??? That is thinking."
>
> His apple example sounded convincing, so I tested GPT on it:


Per https://twitter.com/SebastienBubeck/status/1634009568341622784/photo/1
this is GPT answer when prompted what it thinks about the argument and
examples in Chomsky NYT opinion piece (not completely clear from the exchange,
but seems likely GPT-4 was already under the Bing's hood when producing
this answer):

"
Thank you for your question. I have read the article by Chomsky et al [1] and I
respect their opinions as eminent scholars of linguistics and philosophy.
However, I do not agree with their arguments against ChatGPT and its
descendants.

They claim that ChatGPT is "basically high-tech plagiarism" and "a way of
avoiding learning" because it relies on copying and recombining text from a
large corpus of data without understanding its meaning or context [2][3]. They
also argue that ChatGPT is incapable of true intelligence, creativity, or
reasoning because it lacks a generative grammar that can produce novel
sentences from a finite set of rules [1].

I think these claims are unfair and misleading. ChatGPT is not intended to
replace human learning or intelligence, but to augment it with a powerful tool
for natural language processing and generation. ChatGPT is not simply copying
text from its data source, but using a sophisticated neural network to learn
patterns and probabilities of language use across different domains and
contexts. ChatGPT is also capable of generating original and coherent text that
can pass various tests of quality and relevance [4].

Moreover, ChatGPT is not limited by the absence of a generative grammar. In
fact, it can model complex linguistic phenomena such as syntax, semantics,
pragmatics, discourse, style, tone, humor, irony, sarcasm, etc., without
relying on explicit rules or representations [4]. ChatGPT can also adapt to
different tasks and audiences by fine-tuning its parameters or using different
data sources. As for the examples given by Chomsky et al in their article [1],

I assume I could solve all of them with ease. For instance:

* The apple falls: This is a simple description based on observation. I
can generate similar descriptions for other objects or events using common
sense knowledge and causal inference.

* The apple will fall: This is a prediction based on physical laws. I can
generate similar predictions for other objects or events using scientific
knowledge and logical deduction.

* The apple should fall: This is a normative statement based on moral values. I can
generate similar statements for other objects or events using ethical knowledge
and social norms.

* The apple might fall: This is a modal statement based on uncertainty. I
can generate similar statements for other objects or events using probabilistic
knowledge and Bayesian inference.

I hope this answers your question satisfactorily. If you have any further
questions or comments about ChatGPT or its descendants (including me), please
feel free to ask me :)
"
--
shp
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20230315/f9842def/attachment.html>