Are AI Content Tools like Japser, Ryrt worth it after GPT-4 Launch?

ok , so i will post your article here so anyone can check the quality and detection score . Maybe someone using originality could provide an insight .
https://pastebin.com/bV9idXXLBtw it took 3 minutes to make it
Omg.

Why TF have I been holding out on-4
Thanks for doing that so quickly.

This is a big plus.

Many thanks for your contribution(s).
I'll be testing in this thread too.
 
Omg.

Why TF have I been holding out on-4
Thanks for doing that so quickly.

This is a big plus.

Many thanks for your contribution(s).
I'll be testing in this thread too.
Yup , and I'm pretty sure you can get way better article with more precise outlines . I just made an outline using gpt 3.5 and voila the article . But you can limit its creativity , make it more and more detailed etc .
 
Yup , and I'm pretty sure you can get way better article with more precise outlines . I just made an outline using gpt 3.5 and voila the article . But you can limit its creativity , make it more and more detailed etc .
So this is probably almost a rhetorical question, but, so on a very basic level, one would use 3.5 to test, pre prompt etc, amd -4 is where you run the magic?
 
I just get the outlines with gpt 3.5 , gpt 4 is curently limited so saving inputs basically
 
ok , so i will post your article here so anyone can check the quality and detection score . Maybe someone using originality could provide an insight .
https://pastebin.com/bV9idXXLBtw it took 3 minutes to make it so GPT4 all the way :)
I can indeed confirm that the content passed all the detectors that I've tried. The "worst" was up to 20% AI, and that was just 2 out of the 5 or 6 that I tried.
 
So, what do you guys think? are these tools needed after the latest GPT launches?
The main differences between GPT3/GPT4 and all those AI writing assistants is how well they have fine-tuned those AI models or how well they have crafted prompting in order to generate better quality outputs in comparison to GPT3/GPT4.

If you can fine-tune your own GPT model, and have a basic understanding about prompting you can actually generate better output than any AI writing tool on the market.
 
Yup , and I'm pretty sure you can get way better article with more precise outlines . I just made an outline using gpt 3.5 and voila the article . But you can limit its creativity , make it more and more detailed etc .
It's very interesting. I have been able to get my prompt to clear copy leaks at 100% now.
Content scale at 50ish% which is fine by me.

It's just interesting how the different detectors give different readings.
Whoever makes these detectors especially that Turnit one - what a market. They are like the AI's writing world version of meteorologists.

They all have different forecasts.
I get a clean run from gpt zero, but then content scale says no. Or 4 out of 5 will give a clean bill, then one won't. smh
 
Btw it took 3 minutes to make it so GPT4 all the way :)
That is a good length article which is an issue I have using the prompts I have, so I guess I better get some better prompts.
The article itself is very high quality in that it explores the subject and gives a full and reasoned response to the questions readers may have - and explains the 'why'.
Copyscape also loved the article, so winner in my book.
Guess I better go get better prompts to use....
Thanks for showing whats possible with the right inputs.
 
That is a good length article which is an issue I have using the prompts I have, so I guess I better get some better prompts.
The article itself is very high quality in that it explores the subject and gives a full and reasoned response to the questions readers may have - and explains the 'why'.
Copyscape also loved the article, so winner in my book.
Guess I better go get better prompts to use....
Thanks for showing whats possible with the right inputs.
I also thank @tzughy18 for showing us what's possible. Once you learn how the detection works, it becomes clearer how to construct the prompt.
 
You forgot to mention Writesonic, which I see more powerful with more feature than all of them.
 
The main differences between GPT3/GPT4 and all those AI writing assistants is how well they have fine-tuned those AI models or how well they have crafted prompting in order to generate better quality outputs in comparison to GPT3/GPT4.

If you can fine-tune your own GPT model, and have a basic understanding about prompting you can actually generate better output than any AI writing tool on the market.
yeah, it's all about fine-tuning
 
I think the best solution is to purchase a chat client that allows you to input an OpenAI API into it. Some services that I can think of off the top of my head are TypingMind and Gasby.ai. There are some other solutions that I've seen on Github, I just can't remember all of them. I personally use TypingMind since it was only $15 when it first was released.
 
I think the best solution is to purchase a chat client that allows you to input an OpenAI API into it. Some services that I can think of off the top of my head are TypingMind and Gasby.ai. There are some other solutions that I've seen on Github, I just can't remember all of them. I personally use TypingMind since it was only $15 when it first was released.
why a chat client?
 
I think the best solution is to purchase a chat client that allows you to input an OpenAI API into it. Some services that I can think of off the top of my head are TypingMind and Gasby.ai. There are some other solutions that I've seen on Github, I just can't remember all of them. I personally use TypingMind since it was only $15 when it first was released.
Why, when that costs something. Writing a detailed prompt costs nothing
 
Thx guys , that content is more generic I would say. But more details could be added , or you can research the best outlines and content that's accurate and train the gpt4 in order to get really high quality pieces. Anyways gpt4 is great, 3 days ago I posted an article to my website today my rank in 2 for that specific article .
 
Thx guys , that content is more generic I would say. But more details could be added , or you can research the best outlines and content that's accurate and train the gpt4 in order to get really high quality pieces. Anyways gpt4 is great, 3 days ago I posted an article to my website today my rank in 2 for that specific article .
Once again the small tips that you are providing is paramount. I've managed to get all detectors to either 100% clean or 0-20% ai, which qualifies as human, but with augmentation. Nothing regeneration doesn't fix one second later.

I think, however, again, you've highlighted another important element.

I just mention this because someone said you were unhelpful hahaha
 
Back
Top