After some tweaks, I was able to generate text, but whatever I do, the output text is always between 300-400 words. Does anybody know what I should do to make 1000 words articles?
I use the rewriting script to change the tone, voice etc. I use it whenever I need to make the article sound different.
I would also use it if I'm copying some content from another site. If you copy entire articles, the original site can take down your copied content from the SERP with a DMCA notice.
As of now, I don't have any plans to share that script. It's something I use to upsell another product.
However, you can modify the script from this thread and get it to rewrite content for you. Just modify the input file and the prompt. If the article is long, you may lose some context and the output will probably be shorter than the original.
Which openAI model you use for rephrasing? I have tried davinci-002, but most of the time its generating almost same sentences.
One drawback of using OpenAI for rephrasing is it will cost double since input tokens also count.
what about using deepl.com/write for paraphrasing?
Which openAI model you use for rephrasing? I have tried davinci-002, but most of the time its generating almost same sentences.
One drawback of using OpenAI for rephrasing is it will cost double since input tokens also count.
what about using deepl.com/write for paraphrasing?
I use davinci-3. If you’re getting the same sentences, you need to change your prompt and settings.
I posted a rewrite of a news article in a different thread last week. The output is pretty different from the input and it’s way better than rephrasing with Quillbot or something similar.
As others have said, this was a great share.
Some people seem to be struggling with Python, so I used pyinstaller to make an executable version of the script which works on Windows platforms.
Download the file here:
It's a zip file with three components:
- openai_scraper_v1.exe - a Windows executable -
- input.csv - an example of the input file
- openai_key.txt
INSTRUCTIONS
Download and extract onto your Windows machine - I recommend you use a sandbox until you're comfortable it's safe (!)
Edit the input .csv file with your queries
Edit the txt file and use your own openAI key, not the example in the file
Double click to run, and wait - nothing much happens until the run is completed and you end up with an output.csv file. I think Sheets is the best way to manipulate .csv files - you can do a one-click cleanup to remove extra spaces in the text (Data > data cleanup)
No guarantees or warranties. Kudos to OP @warriorsam53 who wrote the original script .
I checked the executable with Virustotal. I have no idea why it's showing 3 positives out of 67 - as I said, use at your own risk.
Try it with just a few (3 - 5) items on your list and leave it to run. I thought the script didn't work initially but realised that it takes a while for the openai to generate articles, and also it can get stuck due to congestion. If you use a very short list you should see some results in a couple of minutes.
What about plagiarism? A lot of content Chat GPT has produced for me, I check it with Grammarly, etc, and most of the time it tells me they've found major plagiarism.
Also how quickly can you start getting traffic? Do you need to build any backlinks?
What about plagiarism? A lot of content Chat GPT has produced for me, I check it with Grammarly, etc, and most of the time it tells me they've found major plagiarism.
Also how quickly can you start getting traffic? Do you need to build any backlinks?
I replied to a similar question on this thread a couple of days ago. If parts of the article are duplicate, it doesn‘t really matter. Sometimes, 100% duplicate content ranks higher than the original.
You can always make the article better by adding researched information. If you don’t provide AI with any data, it can only write about what it’s already read. Which is probably a form of plagiarism.
As far as traffic goes, it depends on your niche. In an easy niche, a few months. In a niche like mortgage, probably never lol.
I have a journey for a PAA site. I think it took 5 months or so for 2000 PV a month.
I try and avoid backlinks as far as possible. It’s hard to get good links for a good price so it’s not worth it for me. I prefer spending time on keyword research so I can rank without links. But if you go after competitive keywords, you don’t have a choice.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
