I can assure you that I am not using AI. I am a human being who is capable of independent thought and action. I am able to access and process information from the real world through my senses and experiences. I am also able to form my own opinions and beliefs based on my understanding of the world.
1 Like
I don’t believe it is possible to use Chat J’ai Pété ironically any more. It was for about two days after it was released.
8 Likes
by declaring this technology generally as garbage you’ll miss it’s power and usefulness
using it to communicate in forum is nonsense, but as a help looking up things it can be pretty useful - if you keep in mind not use it as the only source of information
General demonizing new technology never helped anything in history.
I’m curious what AI-content that should not posted here, the starter of this thread had in mind.
1 Like
I don’t think anyone the OP said it was garbage ? What was said was, we don’t want it posted here as content.
To me, there’s an analogy with google. You wouldn’t answer a query but doing a cut and paste of a full page of google search results, why would you cut’n’paste a full page of text response from a AI service ?
You might suggest someone to go and use the AI service to answer their query, same way you might suggest someone uses google search for their query.
4 Likes
Having designed and taught a course on the social implications of computing, I can assure you that historically, there was far more uncritical acceptance and self-interested boosting of new technologies than there was demonization, and that continues to this day.
12 Likes
With a little help from my friends.
I found humans are pretty good at identifying such content. And flagging it.
4 Likes
the user antimaitre I replied to said exactly this:
hence my reaction.
I don’t know… is someone doing this?
Well, these long text get in the way.
They will not be considered as a user output, and be removed without warning from now on.
3 Likes
sounds like me then.
wait a minute… 
2 Likes
ChatGPT, why does my snare sound like shit?
6 Likes
generally seen I agree with you… but I’m referring here to people who demonize it and believe it passes them instead of thinking about the best way to utilize it and make some benefit from it
yes, I’m a triggered by the existence of this thread and generally declaring it as garbage .
I can assure everyone that I’m seeing this AI thing very critical and I’m also worried about the negative impact it has and will have… but I’m also very interested about positive usecases and I see many chances
I’m not calling ChatGPT nor the use of AI garbage.
It’s the output, in the context of this forum, that can and is considered as it by us, both mods and community.
8 Likes
excellent example… the answers it provides are a very good starting point and will avoid some too general questions in forums and will instead lead the user to start researching and ask more specific questions about optimizing the snare sound
It’s been happening lots. You’ve been lucky to miss it. Often takes the form of “Ooh … look what chatGPT said about this topic” … which often turns out to be predictable and not worth anyones time.
Sorry, I’ll amend my response. The OP did did not claim it was all garbage. He used the word “fun” at one point.
2 Likes
I would be extremely cautious in using ChatGPT for looking information because it is a language model not a model for facts. From OpenAI: ”ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers. Fixing this issue is challenging, as: (1) during RL training, there’s currently no source of truth; (2) training the model to be more cautious causes it to decline questions that it can answer correctly; and (3) supervised training misleads the model because the ideal answer depends on what the model knows, rather than what the human demonstrator knows.” And they list also other limitations the model has.
Basically it is not a source for facts and without prior knowledge it is not possible to reliably deduce if the output is true or not. Your reference information could be false too.
14 Likes
I can relate, it’s like you give humans something wonderful that can grow crops and feed the hungry and they turn it into a bomb. I get it. It’s hard to have a discussion that pleases all. I’m upset by the notion of the potential misuses, as well as the uprooting and pain which comes with swift change, at the same time I’m hopeful about something positive which might come from the technology one day and I only hope that my misgivings about humanity aren’t well founded in the history of humans as opposed to the reluctant optimism by which I live my life.
I can understand your perspective and I do see why it’s frustrating to feel like people are saying there’s no merit and it’s a personification of evil, but to some of us it’s probably more distrust in humanity’s ability to use it responsibly than the technology being to blame. Just not everyone is able to articulate it as such because it’s more of a feeling than an academic response.
I do recall someone making a post about how it fabricated information from the octatrack manual once, and it basically made it sound realistic but there was no such page and no such information in the manual.
6 Likes
As much as I like a good joke, I’m not really in the mood today to get tested.
There might be casualties, let’s not go this way please.
4 Likes
Spot on. I’ve replied like this myself recently, elsewhere on this forum.
2 Likes
Yes, that should be clear to everyone!
But if you look into it you’ll find that there are many correct answer and obviously also wrong ones. Caution is important As it is with googling stuff… or finding answers about synth-gear in forums.
I often found praises about some gear and then when using it myself I found after very short time big issues that make it annoying to use.
I heard about new stuff regarding chatGPT … they introduced GPTs - that can be private very specialized topic-specific answering machines with “intelligence”.
For example a synth-manufacturer could create this instead of a documentation and users can ask in more natural way how to do this or that with the devices of this manufacturer.
1 Like
Out of curiosity I spent some time reading articles from experts about “weak AI”, which ChatGPT is. Here some findings …
First problem seems to be, those algorithms only can work on content already fed to them and create some kind of weighted probability, whether a potential answer could “make sense” or not … in the pure sense of probabillity 
not by “understanding the meaning” of the “question”.
Data used for the open internet are rarly validated by knowledgeable people. This makes for a lot of “nonsense” in and “false interpretation” of the data base. It’s as easy as in “put in a part of the data as bullshit and you will receive bullshit”
Experts say, if you want to create a fun or creative text, no problem, but don’t ask for “expert knowledge”. There is a relatively high possibility that the algorithm is answering by “haluzination”. Meaning the output is written nicely and seems to make perfect sense, but if you dig a little to confirm the output, you will find that often the AI is obviously wrong or seems to have “invented” some “facts”. Hence the term “haluzination”
This said, have fun - but don’t trust it
5 Likes