hacking 1000 articles.

[ArYit]

really ?

 

[M4DM4X]

can i use the SB premium article plugin to scrape non-english site and translate to english? i mean i dont want to scrape the pre selected article directories but real sites. Is this possible with SB? Thanks

Yep

 

[crnack]

OK so I tried scrapebox and wordai. What a waste of $$ Scrapebox is. It's literally a piece of shit. I might be able to hack with Wordai but it will be a painfully slow process because of how crap the NLP is...
But we got to experiment with these things to find the holes I guess...

I completed 1x 9k article today. It took about 4 hours. Going to try 1-2 more times and then determine whether it's worth pursuing - ie, I can speed it up enough to warrant the investment..
My instinct has me thinking I will need to roll my own NLP with T5, Pegasus, or something...

Cmon mate. You can do it. It's only 999 left.

 

[feralAddict]

Scrapebox is a piece of shit. who gives a shit about keywords.
See my previous posts as to why I don't care about keywords.

Besides, Scrapebox relies on Wordai which is also a pile of shit. And the other Scrapebox extensions are also a pile of shit - yes I forked out money for the premium article scraper - what a waste.
I could have got myself some nice takeaway for that $$.

I want high-quality, valuable content. Not some crap that doesn't make any sense, like the crap you get from Wordai.

Besides that... the problem is now cracked!

I have a working solution to produce 1k articles of any length I want. of HIGH QUALITY, well-written, engaging content that not only will google love but people will also!

What an absolute victory this shit is... took a bit longer than I hoped but the end result is better than anticipated. AND I came in under budget!

Keep an eye on Producthunt in the next couple of weeks players!!!

This thread is done grinners are winners.

PS. My next thread will be on hacking PH... stay tuned.

 

[feralAddict]

try with 1000K you will see !

Yep. My solution once in production can handle this with no problem - we might even try to do the entirety of Medium as a stress test... hmmm. it is tempting. I will need to estimate the compute cost and do a cost/benefit analysis. $$$ vs Klout

 

[feralAddict]

I think he was too lazy and just aimed for a quick buck.

Dude. Hell yeah, I'm lazy. I'm not going to waste my time trying to optimize a shit solution. Everyone should be attacking work using the Pareto principle to guide capital allocation. 80/20 for the win.

If I can't hack a 10x solution on whatever shit is currently being used, then why would I bother. If I did, then guaranteed some other loser would 10x me = waste of my capital.

Or worse, they might only 1x my solution... then I might be stupid enough to 1x them... or even less. Then, it will be a race to the bottom before we know it. And I might even find myself trying to flog SEO services... *cringe*

Optimization is for the [INSERT WHATEVER UNSAVOURY BOTTOM FEEDING ANIMAL COMES TO MIND]!

 

[feralAddict]

Cmon mate. You can do it. It's only 999 left.

Yes. Yes, I can. And I did. I hope you enjoyed this thread, and I hope you get to have a play with the solution once we release on PH or HN or where ever it ends up.
And I also hope you follow my next thread on hacking PH... hmmm... I would also like to try articles > video so I can hack youtube... I might need to use deepfake though which could get compute heavy.

Nah, let's hack PH first then do articles > vids!

 

[feralAddict]

Write 1M Articles Instead of 1000
Just Kiding

I'm confident my solution would be able to handle 1M high-quality articles. If I monetised at $1 per 500 words, I don't think it would take long to hit the 1M article mark. If I deployed on AWS I would get at least $5k worth of credits, plus I could wrangle at least $10k more... I will need to check the compute cost and the demand... although I cracked the solution today and already managed to find a partnering marketing agency... could be a worthwhile little hack.

 

[feralAddict]

Did you use their premium addon "article scraper" ?
http://www.scrapebox.com/article-scraper-pluginI'm surprised you found it useless, this addon is a beast, you can havest articles based on html tags or kw, spin, translate, autopost, etc

nah man. scrapebox is not scalable and wordai is shit quality. I tried a few articles and it is so easy to tell it is spun content. half the time it gets the sentence meaning completely wrong.
a easier to use more scalable solution than scrapebox is to whip up a few lines of BS4 or selenium and then you got your scraper. As for auto-posting, every CMS has an API so it takes a few seconds to set up autoposting.

 

[crnack]

I'm confident my solution would be able to handle 1M high-quality articles. If I monetised at $1 per 500 words, I don't think it would take long to hit the 1M article mark. If I deployed on AWS I would get at least $5k worth of credits, plus I could wrangle at least $10k more... I will need to check the compute cost and the demand... although I cracked the solution today and already managed to find a partnering marketing agency... could be a worthwhile little hack.

If you can offer something like that, I think you could just get some venture capital. That's it.

 

[crnack]

nah man. scrapebox is not scalable and wordai is shit quality. I tried a few articles and it is so easy to tell it is spun content. half the time it gets the sentence meaning completely wrong.
a easier to use more scalable solution than scrapebox is to whip up a few lines of BS4 or selenium and then you got your scraper. As for auto-posting, every CMS has an API so it takes a few seconds to set up autoposting.

So you write a few lines with BS4 or Selenium and what then? Post duplicate content?

 

[feralAddict]

What is the "9k" for here? Is this about the word count? If so, how many keywords have you targeted using this number?
To be honest it would be enough content to write a book or make a whole website.

Currently, I'm aiming for 20 - 30k readable and unique words per day.
I used a combination of 4 different software pieces to make it work. It's more of a process that requires software for each step.

The test article was a 9k word pillar article. And I don't worry about keywords. SEO tricks (inc keywords) are shit. My philosophy is simply to create content people want to read and keep reading, create a great UX (including a fast site) and increase dwell time by planning topic clusters properly. Make sure your CTA's are sensible and high-value lead magnets.

 

[feralAddict]

So you write a few lines with BS4 or Selenium and what then? Post duplicate content?

Feel free to read the thread. I'm producing 100% original content. This is still on topic, informative and accurate - gpt3 can't be depended on to actually write valuable content.

 

[crnack]

The test article was a 9k word pillar article. And I don't worry about keywords. SEO tricks (inc keywords) are shit. My philosophy is simply to create content people want to read and keep reading, create a great UX (including a fast site) and increase dwell time by planning topic clusters properly. Make sure your CTA's are sensible and high-value lead magnets.

People use keywords for promotional purpose. I understand you may believe you will cover the whole industry and by doing that, you will naturally get a lot of keywords in your articles, but it just doesn't seem to work in terms of promotion.

Feel free to read the thread. I'm producing 100% original content. This is still on topic, informative and accurate - gpt3 can't be depended on to actually write valuable content.

I understand you are not using OpenAI API which is trash in my humble opinion.
To put it nicely, content produced by it is like written by a 5 year old who had college level English.

 

[feralAddict]

If you can offer something like that, I think you could just get some venture capital. That's it.

Nah man. VC can fuck off. It's not like winning lotto. There are too many strings. I would rather take my $2m and roll it into another project.

 

[feralAddict]

People use keywords for promotional purpose. I understand you may believe you will cover the whole industry and by doing that, you will naturally get a lot of keywords in your articles, but it just doesn't seem to work.

I understand you are not using OpenAI API which is trash in my humble opinion.
To put it nicely, content produced by it is like written by a 5 year old who had college level English

I don't mind the quality of writing so much. It is the information value that is questionable, IMO. It's a nice model but it's still weak AI and needs to be primed for downstream results.

 

[crnack]

I don't mind the quality of writing so much. It is the information value that is questionable, IMO. It's a nice model but it's still weak AI and needs to be primed for downstream results.

The quality is because the 5 year old understands English and its syntax. But it doesn't understand the broader context. It just doesn't do adult-like research.

 

[feralAddict]

hear hear. It is what happened to me exactly.

Dude I agree the writing is good. However, it's a fact that the information value is low.

 

[india]

1000 articles should make your $$$. Good luck...

 

[WorkingNow]

Though i agree with the 80/20 approach, i disagree on optimization. A 1% improvement compounded daily really adds up. Thats how giants are born, and as the saying goes "big things have small beginnings".