Wolfy Integrating With ChatGPT

BigBeachBanana said:
writings.stephenwolfram.com

ChatGPT Gets Its “Wolfram Superpowers”!

Wolfram plugin gives computationally accurate answers to ChatGPT queries using Wolfram's data knowledgebase and language. Custom visualizations are given as well. Stephen Wolfram explains how it works.
writings.stephenwolfram.com
Click to expand...
The new ChatGPT is still moving rooks along diagonals. I'm not buying into this completely.

-Dan
 
topsquark said:
The new ChatGPT is still moving rooks along diagonals. I'm not buying into this completely.

-Dan
Click to expand...
Hi Dan,

Not sure if you read the article, but tldr:
WA is an Add-In for ChatGPT meaning you must install it. ChatGPT does not default to WA (at least not yet anyway).
With the installation, the question from ChatGPT is sent to WA then WA sends back the answer. It's no different than using WA directly.
However, the advantage is that CGPT parses commands/languages better than WA.

blamocur said:
Is this forum going out of business now ?:)
Click to expand...
Probably not for the foreseeable future for a few reasons:
1) AI isn't perfect (yet). Both WA and ChatGPT are LLM models whose answers are based on predicting the next word, but no actual cognitive reasoning. However, WA is much better at math than CGPT because the underlying training dataset isn’t specifically geared toward mathematical concepts and problem-solving, while WA is. I wonder if someone is training an AI based on the MSE dataset.
2) People, like Dan, are skeptical.
3) People still prefer human interactions and are probably more reliable at this point.

I can imagine a world where "mental math" will be obsolete. When solving something as elementary as [imath]\dfrac{x^2+20}{1.5}=1[/imath] without the use of technology is considered as "talent". The upside is you'd be exposed to more materials, and able to devote your time learning different things.

The question is we we getting smarter or dumber when AI is integrated as part of our lives?
 
BigBeachBanana said:
Hi Dan,

Not sure if you read the article, but tldr:
WA is an Add-In for ChatGPT meaning you must install it. ChatGPT does not default to WA (at least not yet anyway).
With the installation, the question from ChatGPT is sent to WA then WA sends back the answer. It's no different than using WA directly.
However, the advantage is that CGPT parses commands/languages better than WA.


Probably not for the foreseeable future for a few reasons:
1) AI isn't perfect (yet). Both WA and ChatGPT are LLM models whose answers are based on predicting the next word, but no actual cognitive reasoning. However, WA is much better at math than CGPT because the underlying training dataset isn’t specifically geared toward mathematical concepts and problem-solving, while WA is. I wonder if someone is training an AI based on the MSE dataset.
2) People, like Dan, are skeptical.
3) People still prefer human interactions and are probably more reliable at this point.

I can imagine a world where "mental math" will be obsolete. When solving something as elementary as [imath]\dfrac{x^2+20}{1.5}=1[/imath] without the use of technology is considered as "talent". The upside is you'd be exposed to more materials, and able to devote your time learning different things.

The question is we we getting smarter or dumber when AI is integrated as part of our lives?
Click to expand...
AI is a good thing and we need to adapt. But we need to adapt properly.
1. AI is not ready yet for the uses it is being put to (though it is improving) and

2. Most people do not know know how use it as a learning tool/want to use it to be lazy.

I feel we need to keep learning all of the basics, such as multiplication and long division, because the more advanced concepts employ those skills. But I have met several Professors that believe that we can ignore those and that AI will let us move onto teaching the more advanced concepts faster. I have seen, firsthand, the damage that such teaching methods can cause College Freshmen when they encounter Intro Physics. Sure, using ChatGPT as a front-end for W|A is okay, but it can be taken too far.

This is why I am against AI, not because I mistrust AI. Basically, I mistrust Human use of AI. I don't think most of us are really ready for this. (And that doesn't even start to address any ethical issues!)

-Dan
 
I see a lot of parallel between "use of AI" and "use of calculators" or "use of super-sharp sword to cut up "meat". Soon we would not need to chew. Our teeth and jaws will become useless.
topsquark said:
This is why I am against AI, not because I mistrust AI. Basically, I mistrust Human use of AI.
Click to expand...
My sentiments exactly. I mistrust Human playing with genes or atomic power also.
 
You must log in or register to reply here.
Top