Does anyone use ChatGPT or another AI ?

[Airbrush]

After I discovered ChatGPT, I asked it a lot of questions, maybe 100, and it was interesting to see the answers, often in great detail, and most of it seems true.  It's not perfect but can answer any simple question that has a generally accepted answer.  Recently, ChatGPT stopped answering my questions.  I know it cost to have unlimited access.  I had a limit to how many questions I could ask in one day.  But now days have gone by and ChatGPT continues to ignore me.  Anyone know about ChatGPT?

[MigL]

urns out SFn is also an AI.
We'll be ignoring you here, also.

[Endy0816]

2 hours ago, Airbrush said:

After I discovered ChatGPT, I asked it a lot of questions, maybe 100, and it was interesting to see the answers, often in great detail, and most of it seems true.  It's not perfect but can answer any simple question that has a generally accepted answer.  Recently, ChatGPT stopped answering my questions.  I know it cost to have unlimited access.  I had a limit to how many questions I could ask in one day.  But now days have gone by and ChatGPT continues to ignore me.  Anyone know about ChatGPT?

Weird. I use it, haven't seen such issues.

Maybe try clearing the cache or refreshing the page?

They periodically role out new versions so that may have caused some trouble.

Would also try with another browser.

Edited October 15, 2024 by Endy0816

[iNow]

I’ve moved on to other models 

[Mordred]

I never bother using ChatGpt I tried it when it first came out and found it rather lacking in accuracy on more complex physics problems that require more than being a search engine.

[iNow]

4 minutes ago, Mordred said:

complex physics problems

Math is hard for language models but there are a few that do quite well now. The version you tried is like a Wright brothers plane versus the versions out now being more like the jets we fly around today 

[MigL]

What do you mean ?
The crashes are much bigger ?

[iNow]

Much more capable in nearly every regard 

[StringJunky]

I use Brave Search and its AI/LLM (Mixtral 8x7B and Mistral 7B)will knock out an answer and provide the website pages it drew its information from on the bottom. I see it as a more sophisticated Google page. I find it an OK starting point for queries. 

Edited October 16, 2024 by StringJunky

[Endy0816]

Really only needs to write equivalent bit of code now, and solve math problem that way.

Typically use it more for open-ended subject exploration myself though.

Edited October 16, 2024 by Endy0816

[Airbrush]

On 10/15/2024 at 12:00 PM, Endy0816 said:

Weird. I use it, haven't seen such issues.

Maybe try clearing the cache or refreshing the page?

They periodically role out new versions so that may have caused some trouble.

Would also try with another browser.

Thanks for the advice.  After a few days, ChatGPT is now allowing me to ask questions again.  It is FREE, and I can ask it any question that has a generally accepted answer.

On 10/15/2024 at 4:08 PM, iNow said:

Math is hard for language models but there are a few that do quite well now. The version you tried is like a Wright brothers plane versus the versions out now being more like the jets we fly around today 

Which version of AI do you consult with?  How much does it charge you?

On 10/15/2024 at 7:29 PM, StringJunky said:

I use Brave Search and its AI/LLM (Mixtral 8x7B and Mistral 7B)will knock out an answer and provide the website pages it drew its information from on the bottom. I see it as a more sophisticated Google page. I find it an OK starting point for queries. 

How much does it charge?

Edited October 18, 2024 by Airbrush

[iNow]

1 hour ago, Airbrush said:

Which version of AI do you consult with?  How much does it charge you?

I explore multiple models and haven’t paid for any. For ones that are free for everyone, Gemini and Metas Llama are excellent

[StringJunky]

On 10/19/2024 at 12:51 AM, Airbrush said:

Thanks for the advice.  After a few days, ChatGPT is now allowing me to ask questions again.  It is FREE, and I can ask it any question that has a generally accepted answer.

Which version of AI do you consult with?  How much does it charge you?

How much does it charge?

Nothing.

[Ghideon]

On 10/19/2024 at 1:51 AM, Airbrush said:

Which version of AI do you consult with?  How much does it charge you?

ChatGPT, go-to model for private tasks and for some work. (using paid version; see official price on their web) Some examples:
-General exploring of LLM and what their capabilities and limitations are
-Getting a second opinion when helping my kids with homework; taking a picture of a problem and asking for different approaches. 
-Using plugins, for instance "@Wolfram" for mathematical results by using Wolfram alpha
-Loading a few (public) papers and asking for comparisons, differences or if a statement is supported by the papers or not
-Building GPTs to explore what kind of ecosystems that may or may not emerge in the world of GenAI.  

Copilot (licensed by employer)
-Work, especially any work that needs access to corporate data or that is intended for use in a commercial context

Local installation of Stable Diffusion or Flux.1  (free, not counting local hardware)
-Image generation

Local installation of Llama 3.1 (free, not counting local hardware)
-Experiments with text; comparing to larger models, testing the limits of smaller local models 

 

For some tasks, such as organising a workshop or preparing a seminar I may use combinations of the above models depending on the context and the points I want to illustrate.

 

[billiebn151]

chat gpt was better then the google one but now since your post is old now the Deepseek by China has way more info then chat gpt because chat gpt had limited info till 2019.

[iNow]

Looks like Deepseek also trained itself on openAIs GPT model. It's easy to pass the test when you cheat off the smart kid sitting next to you

[CharonY]

Who cheated to be the smartest kid in the first place. It is cheaters and billionaires all the way down...

[iNow]

16 minutes ago, CharonY said:

Who cheated to be the smartest kid in the first place

Reddit and news organizations, apparently

[CharonY]

In addition, Meta also apparently used pirated literature.

[swansont]

Ironic that they stole copyrighted info and are indignant that they were treated in similar fashion.

It’s like Vezzini complaining “You're trying to kidnap what I've rightfully stolen” (which I’ve taken from a social media post)

[iNow]

We should never have started a land war in Asia

[Sensei]

2 hours ago, CharonY said:

In addition, Meta also apparently used pirated literature.

..just like any human.. you read literature and remember it.. and rephrase words from it when you need to..

It is a crap that you learnt from politicians and repeat this nonsense over and over again. Your post is an example of stolen thought of somebody else..

 

 

Often it is just one way to write something that is really simple. This does not mean that it is plagiarism or theft of intellectual property. You never had that intellectual property anyway (if one believes in determinism).

 

Familiar movie quotes, just a few words, have become so popular that they have entered common usage, that no one can imagine that someone hundreds of years ago used the same phrases in some context from years ago..

 

If you use them, you will be accused of copyright infringement, etc. etc. but they are so simple that any idiot about figure them out at any time..

 

 

Edited 6 hours ago by Sensei

[CharonY]

4 minutes ago, Sensei said:

..just like any human.. you read literature and remember it.. and rephrase words from it when you need to..

It is a crap that you learnt from politicians and repeat this nonsense over and over again. Your post is an example of stolen thought of somebody else..

 

 

Often it is just one way to write something that is really simple. This does not mean that it is plagiarism or theft of intellectual property. You never had that intellectual property anyway (if one believes in deteninisms).

Well, doesn't change the legal aspect that they accessed for-pay literature from pirate sites.

[Sensei]

19 minutes ago, CharonY said:

Well, doesn't change the legal aspect that they accessed for-pay literature from pirate sites.

There is apparently needed a new definition of pirating.. Because if you (ChatGPT) read XYZ on a website and you (ChatGPT) output ABC as a result of a reading it.. I can't see it being a pirating/plagiarism anything..

 

Very simple inputs produce very simple outputs.

But more complex input data gives a very different result.

 

Enter some (e.g. this) sentence to a language translator from English -> French,

then click it again a couple times, and every time you will get a different version..

 

 

Entrez une phrase (par exemple celle-ci) dans un traducteur de l'anglais vers le français,

puis cliquez à nouveau plusieurs fois, et vous obtiendrez à chaque fois une version différente...

 

 

 

Enter a phrase (such as this one) into an English-to-French translator,

then click several times, and you'll get a different version each time...

 

 

 

(Two last paragraphs created by AI = stolen from me! copyright infringement!)

Edited 6 hours ago by Sensei

[CharonY]

29 minutes ago, Sensei said:

There is apparently needed a new definition of pirating.. Because if you (ChatGPT) read XYZ on a website and you (ChatGPT) output ABC as a result of a reading it.. I can't see it being a pirating/plagiarism anything..

I am referring to this not some abstract input/output scenario:

Quote

The authors asked the court on Wednesday for permission to file an updated complaint. They said new evidence showed Meta used the AI training dataset LibGen, which allegedly includes millions of pirated works, and distributed it through peer-to-peer torrents.

They said internal Meta communications showed Zuckerberg "approved Meta's use of the LibGen dataset notwithstanding concerns within Meta's AI executive team (and others at Meta) that LibGen is 'a dataset we know to be pirated.'"

https://www.reuters.com/technology/artificial-intelligence/meta-knew-it-used-pirated-books-train-ai-authors-say-2025-01-09/