Welcome to the staging ground for new communities! Each proposal has a description in the "Descriptions" category and a body of questions and answers in "Incubator Q&A". You can ask questions (and get answers, we hope!) right away, and start new proposals.
Are you here to participate in a specific proposal? Click on the proposal tag (with the dark outline) to see only posts about that proposal and not all of the others that are in progress. Tags are at the bottom of each post.
Post History
I created the AI tech proposal primarily because I thought it would be good to have a place to discuss doing stuff with AI models rather than the details of how they are trained and how they work. ...
#2: Post edited
by
ArtOfCode
·
2023-08-02T18:32:07Z (over 1 year ago)
- I created the [AI tech proposal](https://proposals.codidact.com/posts/289124) primarily because I thought it would be good to have a place to discuss *doing stuff with* AI models rather than the details of how they are trained and how they work. Also, LLMs and prompt engineering appear to have become a bit of an odd duck in technical discussions. At the same time I think these are interesting topic and cool things can be accomplished with them if used *correctly*.
- Funnily enough, StackExchange launched [GenAI](https://genai.stackexchange.com/) shortly before I created the proposal - although I didn't know about it at the time. The SE site of course has more activity, but I think Codidact is inherently able to tackle this area better, because it has a more flexible moderation approach.
- However, on reflection, I'm not so convinced that it's worth focusing the site on just large models. For one, the distinction is kind of blurry. Two, it wouldn't really *hurt* if people were *also* talking about smaller models on the same site. It might even help, because someone primarily interested in small models may have helpful insights into large models as well - after all they are not that different other than scale. Lastly, narrow scope is something that helps with the "too much content" problem, but if anything we have a "too little content" problem. It would be better to start with a broadly-scoped site, and split it later when (and if) an issue actually arises.
I see that there is also a (Machine Learning proposal)[https://proposals.codidact.com/posts/289179]. I think that one is actually better scoped for a section that encompasses both LLMs as well as the rest of ML/AI. However, currently "prompt engineering" is off topic for the ML proposal - I wonder if there's any room for reconsidering that.- * Does it make more sense to have a general, broadly scoped section about ML or have several more specialized ones?
- * Should the AI tech and ML proposals be combined? (I'm happy to remove my AI tech proposal if the scope of the ML one can expand a bit)
- * For incubator questions that potentially fit both proposals, would it be reasonable to tag them with both?
- I created the [AI tech proposal](https://proposals.codidact.com/posts/289124) primarily because I thought it would be good to have a place to discuss *doing stuff with* AI models rather than the details of how they are trained and how they work. Also, LLMs and prompt engineering appear to have become a bit of an odd duck in technical discussions. At the same time I think these are interesting topic and cool things can be accomplished with them if used *correctly*.
- Funnily enough, StackExchange launched [GenAI](https://genai.stackexchange.com/) shortly before I created the proposal - although I didn't know about it at the time. The SE site of course has more activity, but I think Codidact is inherently able to tackle this area better, because it has a more flexible moderation approach.
- However, on reflection, I'm not so convinced that it's worth focusing the site on just large models. For one, the distinction is kind of blurry. Two, it wouldn't really *hurt* if people were *also* talking about smaller models on the same site. It might even help, because someone primarily interested in small models may have helpful insights into large models as well - after all they are not that different other than scale. Lastly, narrow scope is something that helps with the "too much content" problem, but if anything we have a "too little content" problem. It would be better to start with a broadly-scoped site, and split it later when (and if) an issue actually arises.
- I see that there is also a [Machine Learning proposal](https://proposals.codidact.com/posts/289179). I think that one is actually better scoped for a section that encompasses both LLMs as well as the rest of ML/AI. However, currently "prompt engineering" is off topic for the ML proposal - I wonder if there's any room for reconsidering that.
- * Does it make more sense to have a general, broadly scoped section about ML or have several more specialized ones?
- * Should the AI tech and ML proposals be combined? (I'm happy to remove my AI tech proposal if the scope of the ML one can expand a bit)
- * For incubator questions that potentially fit both proposals, would it be reasonable to tag them with both?
#1: Initial revision
by
matthewsnyder
·
2023-08-02T18:17:43Z (over 1 year ago)
Broaden the scope of AI tech into all ML?
I created the [AI tech proposal](https://proposals.codidact.com/posts/289124) primarily because I thought it would be good to have a place to discuss *doing stuff with* AI models rather than the details of how they are trained and how they work. Also, LLMs and prompt engineering appear to have become a bit of an odd duck in technical discussions. At the same time I think these are interesting topic and cool things can be accomplished with them if used *correctly*. Funnily enough, StackExchange launched [GenAI](https://genai.stackexchange.com/) shortly before I created the proposal - although I didn't know about it at the time. The SE site of course has more activity, but I think Codidact is inherently able to tackle this area better, because it has a more flexible moderation approach. However, on reflection, I'm not so convinced that it's worth focusing the site on just large models. For one, the distinction is kind of blurry. Two, it wouldn't really *hurt* if people were *also* talking about smaller models on the same site. It might even help, because someone primarily interested in small models may have helpful insights into large models as well - after all they are not that different other than scale. Lastly, narrow scope is something that helps with the "too much content" problem, but if anything we have a "too little content" problem. It would be better to start with a broadly-scoped site, and split it later when (and if) an issue actually arises. I see that there is also a (Machine Learning proposal)[https://proposals.codidact.com/posts/289179]. I think that one is actually better scoped for a section that encompasses both LLMs as well as the rest of ML/AI. However, currently "prompt engineering" is off topic for the ML proposal - I wonder if there's any room for reconsidering that. * Does it make more sense to have a general, broadly scoped section about ML or have several more specialized ones? * Should the AI tech and ML proposals be combined? (I'm happy to remove my AI tech proposal if the scope of the ML one can expand a bit) * For incubator questions that potentially fit both proposals, would it be reasonable to tag them with both?