Combining several texts into one

Alpha1404

Senior Member
Jr. VIP
Joined
Mar 21, 2021
Messages
1,032
Reaction score
833
Hello,
with an artificial intelligence tool, is it possible to combine several articles into one, with an intelligent structure, without repetition?
I've tried with ChatGPT, but it gives me a pretty mediocre summary.
ZimmWriter doesn't have the option.
If you have an idea, I'd love to hear from you :)
Thanks
 
Hello,
with an artificial intelligence tool, is it possible to combine several articles into one, with an intelligent structure, without repetition?
I've tried with ChatGPT, but it gives me a pretty mediocre summary.
ZimmWriter doesn't have the option.
If you have an idea, I'd love to hear from you :)
Thanks
Here's one way you can do it.
1) extract all entities and Lsi words from the articles. There must be a tool/chrome extension that can do this. May be chatgpt+ can do this.
2) you can add all of them in the training box in zimmwriter 1 click seo article generator.
3) generate the article.
 
Here's one way you can do it.
1) extract all entities and Lsi words from the articles. There must be a tool/chrome extension that can do this. May be chatgpt+ can do this.
2) you can add all of them in the training box in zimmwriter 1 click seo article generator.
3) generate the article.
Hi there,
yes, that could be an option.
In fact, I have a good hundred articles in a development WordPress. Each article is about a hundred words long.
The idea was to combine them all (well, not all, but you get the general idea) into a coherent whole.
 
what do you mean combine articles? how big is the prompt you're sending? maybe you're running into an issue because you're feeding it hundreds of articles and overflowing the context? Google gemini has a bigger context but you might need to rethink how you're using an llm if you're feeding it a huge amount of text. maybe notebookLM?
 
what do you mean combine articles? how big is the prompt you're sending? maybe you're running into an issue because you're feeding it hundreds of articles and overflowing the context? Google gemini has a bigger context but you might need to rethink how you're using an llm if you're feeding it a huge amount of text. maybe notebookLM?
Make an intelligent article out of two very short articles.
ChatGPT just makes me an assembly of text 1 and 2, without putting stuff around it.
 
Make an intelligent article out of two very short articles.
ChatGPT just makes me an assembly of text 1 and 2, without putting stuff around it.
If it's just feeding in two short articles then I'm surprised ChatGPT isn't working for you. Maybe you need to adjust your prompt to something like "create a new article based on the information from these two articles" rather than use language like "combine these articles" ?
 
Make an intelligent article out of two very short articles.
ChatGPT just makes me an assembly of text 1 and 2, without putting stuff around it.
If that happens I think you need to work on your prompting, bonus: you can ask chat gpt to help you write or improve a prompt. Also sometimes you need a workflow of multiple prompts chained together to get the desired output, Ive done that with the API before to write articles
 
If it's just feeding in two short articles then I'm surprised ChatGPT isn't working for you. Maybe you need to adjust your prompt to something like "create a new article based on the information from these two articles" rather than use language like "combine these articles" ?
If that happens I think you need to work on your prompting, bonus: you can ask chat gpt to help you write or improve a prompt. Also sometimes you need a workflow of multiple prompts chained together to get the desired output, Ive done that with the API before to write articles
I'm giving you both the same answer, because even if your comments aren't identical, the answer is: ChatGPT is totally making stuff up, even when I give him a well-calibrated prompt.
For example, I talk about an almost deserted area, and he writes me a text talking about the many shops, the abundant public transport, etc., etc., etc. In short, he's totally delirious.
 
Well, what GPT does well is rewriting, and whilst it may require some human-made changes in the end, the result was quite impressive for me anytime I used it

There isn't much difference between different LLMs, be they 'writing-specified' or not, so I'd advice promt "rewrite those 2 articles in the same style X and unite them into a single article". Oh, and naming the sections also helps. Like "... into a single article with this, this, and that sections/paragraphs"
 
Well, what GPT does well is rewriting, and whilst it may require some human-made changes in the end, the result was quite impressive for me anytime I used it

There isn't much difference between different LLMs, be they 'writing-specified' or not, so I'd advice promt "rewrite those 2 articles in the same style X and unite them into a single article". Oh, and naming the sections also helps. Like "... into a single article with this, this, and that sections/paragraphs"
Ok, I'll try again :)
 
I'm giving you both the same answer, because even if your comments aren't identical, the answer is: ChatGPT is totally making stuff up, even when I give him a well-calibrated prompt.
For example, I talk about an almost deserted area, and he writes me a text talking about the many shops, the abundant public transport, etc., etc., etc. In short, he's totally delirious.
You're right about ChatGPT sometimes making stuff up, it can be very hard to keep it in check.

While I can't exactly solve your problem, your post inspired me to share some stuff I did a while ago with AI article generation.
I think it gives some ideas to what is possible and maybe could at least give you some inspiration to solving your original problem:
[Guide] - Article generation through a prompt-pipeline
 
Last edited:
You're right about ChatGPT sometimes making stuff up, it can be very hard to keep it in check.

While I can't exactly solve your problem, your post inspired me to share some stuff I did a while ago with AI article generation.
I think it gives some ideas to what is possible and maybe could at least give you some inspiration to solving your original problem:
[Guide] - Article generation through a prompt-pipeline
Thanks : it's always useful :)
 
Back
Top