Using ChatGPT for content?

[PaulEchere]

I see a lot of people refer to GPT as a tool to generate content ideas, which is something I haven't used it for myself.
Just out of curiosity, could someone share some examples of how it helps(ed) to generate content ideas?

[chaumi]

Here's one you can try. It could be improved, but should give you a good idea of what it can give you.....

You are an expert content writer who writes long-form articles. You have a confident tone of voice and are an expert in all forms of online gambling. Your writing is unbiased, unique, and innovative. Create the outline for a long-form English article named "xyzxyzxyzxyzxyx." The list of headings should contain a minimum of 12 separate headings and sub-headings. The outline should be extensive and cover the topic of xyzxyzxyz in depth.

If you can't get it to work, PM me and we can do a screen share.

[dannyx]

There's tools out there that are pretty good at detecting it: https://copyleaks.com/ai-content-detector

It works well in English.
I tested other languages in e.g. in Spanish it is a lottery, a great text will recognize AI and AI will recognize human text, while in Portuguese almost as it flies everything recognizes human text.

[Topboss]

I tried it when it started to become trendy and found it great for generating ideas but not for creating content. At least not to use it directly.

ChatGPT builds its answers from what it finds on the web. It can give great results, but it is sometimes quite inaccurate. Personally, I wouldn't take its answers and copy them into my blog without giving it a thorough review first.

I have the same opinion. It is good for suggesting topics and starting off a sentence if one has momentary writers' block, but in terms of writing accurate content, it is not very good and actually writes a load of rubbish sometimes that sounds credible. I have also checked the content on Copyscape and 50% of the time it has copied stuff word for word from other sites!

When I tried it out, I found myself spending more time correcting it and checking the facts, so it is not worth it as a content writing tool in my opinion, but good for other things like converting a bunch of text into html or adding headings etc.

However one part I found was funny is that I wrote a bunch of text myself and checked it on one of the AI detector tools...and it told me my own writing was AI generated!! I guess I should see that as a compliment lol as that tells me my english must be good!!!

I think it could be a very helpful tool but still has lots of bugs to iron out to be considered accurate.

Asides from all of this, I have clients who pay me to write content for them and I feel that if I use ChatGPT then in essence I would be cheating them as they are paying for human content, not something they can churn out themselves!!

[iso2009]

I think it depends on the niche and on the content you want to create. I will use their API for news and other text for sports betting - still working on the system. First results look amazing, but it's important to know what prompts to use.

Not concerned at all about Google banning it. Yes, it can be easily detected, but that doesn't mean it's useless. Google will be relying more on links, bounces, time spent on site, branded searches and all those other signals.

Already going on full automation with the main project, I will definitely build every new site on auto mode, as much as I can.

[Jamie - RateMyCasino]

It has a lot of bugs and at times writes incorrect content.

[chaumi]

Already going on full automation with the main project, I will definitely build every new site on auto mode, as much as I can.

I think anyone trying to put you off this path would (at this point) be doing you (or anyone taking the same route) a disservice.

But....

At least go in with eyes open and think through what might happen (if you haven't already). I'd suggest it's likely to be a ruinous course for anyone that wants a site that has longevity, at least for anyone that's throwing up dozens or hundreds of pages a day.

But I could be wrong, of course.

[ChrisB]

I'm trying to figure out whether they really believe AI is dangerous or if it's "we are late we need a pause to catch up".

Just like UIGEA.

[ChrisB]

It has a lot of bugs and at times writes incorrect content.

And I think this could be a real problem. I've noticed they (Chat GPT and others) get facts wrong quite often. However, if 'content producers' use these tools frequently, those wrong-facts will be reporduced in many places, which then perpetuates the wrong information to other AI's.

I've seen this in my geneology research on Ancestry. There is an incorrect interpretation of my own ancestors, but so many people have simply copied other people's trees without performing their own research (by which I mean actual birth, death and marriage certificates or church records, not just an Ancestry search), that if you attempt to find the truth, you are swamped with so many trees featuring the wrong interpretation. The result is that most new 'researchers' simply accept the wrong information and reproduce that instead of the correct information.

[iso2009]

I think anyone trying to put you off this path would (at this point) be doing you (or anyone taking the same route) a disservice.

But....

At least go in with eyes open and think through what might happen (if you haven't already). I'd suggest it's likely to be a ruinous course for anyone that wants a site that has longevity, at least for anyone that's throwing up dozens or hundreds of pages a day.

But I could be wrong, of course.

I think GPT is very good at some tasks, while it is awful at others. You need to simply adjust it to what you want. I would not use it to generate the whole content, but only bits and pieces - and mostly to elaborate more on a given prompt. It saves a lot of time and some money.

I just see it as a tool to make some tasks quicker and cheaper, that's all. But full automation (where possible) is, for me, the way to go in the future, and GPT will surely help (as a tool). Mostly for writing/rewriting small pieces of text, and not for using its responses for general inquiries. So here's a big difference between GPT as a knowledge tool (which is dangerous) and using it as a simple tool for writing/rewriting (it does a good job and it's cheaper than any alternative).

As a side note, I would NOT use it to get information for myself, regarding any topic.

[xecutable]

And I think this could be a real problem. I've noticed they (Chat GPT and others) get facts wrong quite often. However, if 'content producers' use these tools frequently, those wrong-facts will be reporduced in many places, which then perpetuates the wrong information to other AI's.

I've seen this in my geneology research on Ancestry. There is an incorrect interpretation of my own ancestors, but so many people have simply copied other people's trees without performing their own research (by which I mean actual birth, death and marriage certificates or church records, not just an Ancestry search), that if you attempt to find the truth, you are swamped with so many trees featuring the wrong interpretation. The result is that most new 'researchers' simply accept the wrong information and reproduce that instead of the correct information.

It even gets simple math wrong, despite writing the correct math formula, showing the correct order of calculating and then coming to the wrong conclusion. The other day, I asked it how would would it take for someone with a 70,000$ net monthly income to become a millionaire. Now I knew it was going to be a year and something, but I wanted to see how many months/days specifically.

ChatGPT's answer was 14 years, 3 months, despite first telling me that 70,000 x 12 will be 840,000 a year, showed me the formula and the calculations and then all of the sudden concluded 14 years, assuming all of the sudden that we are talking about 70,000 yearly.

I told it, it's wrong and should check it again, it apologized corrected itself and when I ask the same question now it answers properly. Point is if something so concrete and factual is so easily mistaken by the AI itself, how can you trust it for far more complex things? However I do find it useful when troubleshooting problems as it give you probably causes, you could have missed.

[Strider1973]

Yes it's not really good a math, did a few tests.
But it's really great in understanding language, it even understands sarcasm! It's able to explain the real meaning of something like this:

Two friends have arranged to study together. While one diligently works on her tasks, the other spends all her time on her phone. Consequently, the diligent one says to her friend:
"Don't forget to take a break once in a while!"

[pgmath]

ChatGPT will never replace the human writing. It deals with words and syntax as information and not with concepts and intentions. AI does not and will never have conceptual thinking. The gambling/casino topics range within a very limited lexical domain compared to other niches. A widespread use the of the robot will ultimately lead to duplicates and no SEO gain.

[Strider1973]

ChatGPT will never replace the human writing. It deals with words and syntax as information and not with concepts and intentions. AI does not and will never have conceptual thinking. The gambling/casino topics range within a very limited lexical domain compared to other niches. A widespread use the of the robot will ultimately lead to duplicates and no SEO gain.

Sorry to say that, but this sounds like you either
- earn your money as a writer
- have never tested ChatGPT
- cannot think a few steps ahead*



* don't forget that in the early days of internet, a fax machine was better suited to transmit an image than an email. And who would have thought that we will watch movies over the internet? What we see here is only the beginning. In 10 years you might get a deepfake video call from your sister, with your sister's voice, to ask you for money... So I'm not saying all will get better, but that's where we are heading to...

[pgmath]

Sorry to say that, but this sounds like you either
- earn your money as a writer
- have never tested ChatGPT
- cannot think a few steps ahead*



* don't forget that in the early days of internet, a fax machine was better suited to transmit an image than an email. And who would have thought that we will watch movies over the internet? What we see here is only the beginning. In 10 years you might get a deepfake video call from your sister, with your sister's voice, to ask you for money... So I'm not saying all will get better, but that's where we are heading to...

Yes, I earn my money as a writer, but this has nothing to do with what I said. Besides, I have no worry that ChatGPT will replace me. The principle stands: AI will never have have conceptual thinking, since this is not algorithmic. In 10 or 1000 years. Your fax example cannot deny that. You may use ChatGPT to fill your page with information (as words connected by syntax and with an one-level semantics), but this will not advance any new ideas, nor connect concepts, nor even arguments. Let aside other predicted issues like duplicates. Trust me.

[wonderpunter]

I done some testing myself.. in the end i deleted it all and kept with manual writing

[dannyx]

I also ran tests but did not upload anything to the site. I personally think that these ai texts were not good.
However, I am sure they will be better than the human ones in the future.
The same as with the translators. I remember X years back for fun one would type something into the translator and it would return silly things. Today, some translation systems are far superior to human translations.

[pgmath]

AI translation is entirely different from article writing. AI translation works because natural language is actually machine-learning-like. Writing articles means conceptualizing, linking concepts (not only words) and expressing intentions, what the AI will never do.

[dannyx]

AI translation is entirely different from article writing. AI translation works because natural language is actually machine-learning-like. Writing articles means conceptualizing, linking concepts (not only words) and expressing intentions, what the AI will never do.

The word never is strong. Surely, if one were to ask a person in 1600 whether it is possible to transport from, for example, Portugal from Latin America in a few/some hours such a person would also say that it would never be possible.

[PROFRBcom]

Like a lot of things it depends what your ultimate goal is. You can use AI generated content and get traffic for sure and right now there is no big penalty for that but we know in the future (if not already) the search engines will be able to spot AI content and the question is what effect that might have.

It's probably fine for churn and burn projects but I would be cautious integrating it into long standing money sites. Then again I am always cautious with new stuff like this and it will be interesting to see how AI content becomes integrated in iGaming and how that will change things.

I agree with your sentiment entirely.

If one does not care about the website's long-term viability, pump out that AI content. I am certain I have seen people doing exactly this in the last few months. Sites I have never seen before jumping into the rankings on page 1... However, I don't think they are lasting long there, not from the handful of times I think I've seen this happen.

For a long-term money producing website, **** putting AI written content on that! That would be too risky.

What I suspect will happen though is that the black-hat community will love this AI ****. Got a PBN? Well, it just became 10x easier to stuff it with "content" to make it look legitimate.

Agree; ChatGPT is a great tool to research and provide content and ideas to generate evening more and better content. Definitely would re-write what ChatGPT returns to better fit your style and SEO requirements.

Next will be for Google to determine ChatGBT content and penalize it.

Ultimately I hope Google detects and penalizes all the AI written content. Ultimately AI content is an echo chamber of **** on the Internet and will end up perpetuating falsehoods, misinformation, and outwrite bullshit as truth and that is dangerous.