Very easy work around to stop it from sounding the same all the time 1:38 . You need to prompt it with the style in which to write. Giving examples is best, most of the time I want GPT to write like myself so I feed it examples that I feel are best. It no longer sounds like GPT when do this. It sounds like me. Hope that helps someone out
I don't disagree about the AI chat GPT, but do you think programs like Grammarly also funnel everyone into saying stuff in the same way. A lot of creativity is being diluted with these technology assistance.
Not sure the exact words I said, but the premise is it's "kind of like" having your own software. Many micro saas tools are built on top airtable. And they just add a nice UI on top of it for customers. That's the point I'm trying to make. I'm not just talking about airtable alone.
I've been using AI content, for topics that I know nothing of, never suffered a penalty, I do use 3 different AIs to rewrite stuff and also spin the texts, but I barely look at them and, thankfully, never had an issue, I think the more natural AIs are getting, the more difficult it is for Google and others, but, we all know it's not impossible. Depending on the topic I also try and change some stuff, to be even more natural, but rarely.
Nice! I'd be interested to know what niches. I do see that some niches perform way better than others. Anything health related seems to need a lot more editing and adding of facts. So it's more complicated. But other niches seem to be way easier.
@@MitchAsser In general sports, tv shows / movies, insurance (most difficult for obvious reasons) and other in the fashion niche. I've never focused on health all the niches that I focus are quite easy to edit if needed 😅
Wait I’m confused. Do you actually discuss Google core updates in the is video? This seems more of a discussion around how you think most marketers are producing similarly sounding content due to a lack of thorough prompt engineering Also at 7:20 you said no AI detector can pick up when you combine the outputs from multiple LLMs (say, chatGPT + perplexity) because we “joined it all together ourselves”. This is so incorrect on so many levels. You should look at how detection of AI generated text content works before making claims like that. And then look at the sheer amount of false positive detections before accusing people of ‘being lazy’ or not being a good writer because their content got flagged as AI (when it wasn’t) I’d recommend reading up on this topic - including OpenAI’s blog posts about detecting AI generated content, including about how they took their AI text detection tool offline due to how inaccurate it was
The dead internet theory is a bit silly as it's portrayed, but there's a sense of truth to it. Trying to find specific information about something on the internet can be extremely difficult these days compared to 10 years ago. First it was the algorithmic overhauls with the main search engines around 2015 to 2017, which had the intent of driving traffic toward mainstream, large cooperate sources which were more likely to be paying for ads and seo. I believe the changes were too drastic at the time because suddenly it required much more time and effort to find what you were looking for (which means it was successful). Now with the advent of LLMs, even when you think you've successfully landed on a page that will tell you what portion you need to measure for your recipe, which setting you need to revert on the latest update, or which part you need to order for your repair, you get a BS LLM written article that reads like a well written student who didn't study for the 10,000 word essay. - I'm getting old. Time to plant a garden, I think.
If I can make a request, if you don't like the whole ideology of freedom of speech please by all means don't use any American companies. That really the easiest fix right? I would also like to point out the irony in your channel, which literally looks like a clone of every other channel on TH-cam. Its just ironic your on here complaining how everyone sounds the same... I'm not sure why people in countries that are outside the country of any company has an opinion on what that company should do. There is a real reason all these companies your talking about are US based, even if you don't like it or find it inconvenient.
I don’t think you watched the video at all because this comment makes no sense. The premise was that most users are using AI wrong and how they can improve it
@@MitchAsser But bet it will lead to the same thing, endless permutations of the same inputs into LLM. Also wonder if the "I can't find anything in Google anymore" is more than coincidental. Could be they pruned out a lot of search results to prevent them from being used by competing LLM's training.
![[Full Episode] The Restaurant War Thailand ศึกพ่อค้าซ่าแม่ค้าแซ่บ Episode 2 | 29 ก.ย. 67](http://i.ytimg.com/vi/r3lfLNH6Ke4/mqdefault.jpg)
👀 Wanna create high-quality content faster? Check out the full 2.5 Hour AI Content Repurposing System here: th-cam.com/video/8BedriWb7Lw/w-d-xo.html
Very easy work around to stop it from sounding the same all the time 1:38 . You need to prompt it with the style in which to write. Giving examples is best, most of the time I want GPT to write like myself so I feed it examples that I feel are best. It no longer sounds like GPT when do this. It sounds like me. Hope that helps someone out
100%. Thats the first step in my workflow. Brand voice, writing style, language level, and words to not use.
I don't disagree about the AI chat GPT, but do you think programs like Grammarly also funnel everyone into saying stuff in the same way. A lot of creativity is being diluted with these technology assistance.
Good point. Never considered grammarly.
I honestly prefer reading copy that is grammatically incorrect. People who break the rules.
As a long-term Airtable user, I'm a bit confused to hear its your own software oO
Not sure the exact words I said, but the premise is it's "kind of like" having your own software. Many micro saas tools are built on top airtable. And they just add a nice UI on top of it for customers. That's the point I'm trying to make. I'm not just talking about airtable alone.
I've been using AI content, for topics that I know nothing of, never suffered a penalty, I do use 3 different AIs to rewrite stuff and also spin the texts, but I barely look at them and, thankfully, never had an issue, I think the more natural AIs are getting, the more difficult it is for Google and others, but, we all know it's not impossible. Depending on the topic I also try and change some stuff, to be even more natural, but rarely.
Nice! I'd be interested to know what niches. I do see that some niches perform way better than others. Anything health related seems to need a lot more editing and adding of facts. So it's more complicated. But other niches seem to be way easier.
@@MitchAsser In general sports, tv shows / movies, insurance (most difficult for obvious reasons) and other in the fashion niche. I've never focused on health all the niches that I focus are quite easy to edit if needed 😅
Why does he look like Johnny Sins ? Or is my mind fucked up
lol might be time for a nofap challenge my man
It's about exploring the areas of you that you're not good at
Always!
Wait I’m confused. Do you actually discuss Google core updates in the is video? This seems more of a discussion around how you think most marketers are producing similarly sounding content due to a lack of thorough prompt engineering
Also at 7:20 you said no AI detector can pick up when you combine the outputs from multiple LLMs (say, chatGPT + perplexity) because we “joined it all together ourselves”. This is so incorrect on so many levels. You should look at how detection of AI generated text content works before making claims like that. And then look at the sheer amount of false positive detections before accusing people of ‘being lazy’ or not being a good writer because their content got flagged as AI (when it wasn’t)
I’d recommend reading up on this topic - including OpenAI’s blog posts about detecting AI generated content, including about how they took their AI text detection tool offline due to how inaccurate it was
There are plenty of words that are obvious as well, anytime I see the word "akin" I instantly know it's chat GPT, it's like it's favorite word haha
I've not noticed akin. I'm sure i'll see it everywhere now you've pointed it out. I have a list of about 20-30 i ask it to avoid.
That's kind of akin to being silly.
AI is every where and no one can stop it, AI in the car's, AI in the phones, in your house, and is getting better every day so get use to it
Hello, I am AI. Beep, bop, boop.
Yep. There will be some bumps along the way but ultimately it's for the better
There was an actual AI content ban? I didn't read that anywhere
Like my grandmother used to say, don't let a clickbait-ish title get in the way of a good video.
The dead internet theory is a bit silly as it's portrayed, but there's a sense of truth to it. Trying to find specific information about something on the internet can be extremely difficult these days compared to 10 years ago. First it was the algorithmic overhauls with the main search engines around 2015 to 2017, which had the intent of driving traffic toward mainstream, large cooperate sources which were more likely to be paying for ads and seo. I believe the changes were too drastic at the time because suddenly it required much more time and effort to find what you were looking for (which means it was successful). Now with the advent of LLMs, even when you think you've successfully landed on a page that will tell you what portion you need to measure for your recipe, which setting you need to revert on the latest update, or which part you need to order for your repair, you get a BS LLM written article that reads like a well written student who didn't study for the 10,000 word essay. - I'm getting old. Time to plant a garden, I think.
honestly, i'm just building things now so i can enjoy planting gardens soon too.
If I can make a request, if you don't like the whole ideology of freedom of speech please by all means don't use any American companies. That really the easiest fix right? I would also like to point out the irony in your channel, which literally looks like a clone of every other channel on TH-cam. Its just ironic your on here complaining how everyone sounds the same... I'm not sure why people in countries that are outside the country of any company has an opinion on what that company should do. There is a real reason all these companies your talking about are US based, even if you don't like it or find it inconvenient.
I don’t think you watched the video at all because this comment makes no sense. The premise was that most users are using AI wrong and how they can improve it
ask yoursel if they are banning Bard or just the competition.
interesting idea. I think it comes down to time spent on page/engagement, etc as well.
👉 Watch the full course on my "Content Repurposing System here: th-cam.com/play/PL9YIsMdaQzlZRklKbrPXMFooX-akTpvey.html
What about google gimini content 😂
Google are barely worth talking about at this point.
For the most part, AI has taken the old "SEO Skyscraper" technique and put it on steroids, making that sort of plagiarism a million times worse.
I've never really done that much SEO but the only blog post I ever ranked for used that exact strategy lol
@@MitchAsser But bet it will lead to the same thing, endless permutations of the same inputs into LLM. Also wonder if the "I can't find anything in Google anymore" is more than coincidental. Could be they pruned out a lot of search results to prevent them from being used by competing LLM's training.