Is STH Shooting Itself in the Foot?

[qwerty27]

Hi Guys, I really like STH and appreciate the thorough writing. However, I noticed today that Patrick used screenshots of AI responses from Google in the article here - https://www.servethehome.com/pickin...n-2025-is-still-complex-amd-oracle-microsoft/

I see this as condoning the stealing of content from other writers and giving them zero credit for the effort, while Google makes money from it. Platforms like STH don't survive by giving content to others who then make money from it, returning nothing to the platforms.

In other words, I don't think STH would copy and paste information from other sites, giving them no credit as the source and no link to the source. Posting a screenshot of these AI responses is essentially that.

 

[rtech]

AI generated content is uncopyrightable AFAIK.

 

[louie1961]

don't think STH would copy and paste information from other sites, giving them no credit as the source and no link to the source. Posting a screenshot of these AI responses is essentially that.

You are wrong on multiple counts. First, they did give credit in the caption. Second, AI text outputs are no copyright protected

"The U.S. Copyright Office released its January 2025 report to address the legal and policy issues related to artificial intelligence (AI) and copyright, as outlined in the Office’s August 2023 Notice of Inquiry. This report has clarified that outputs generated by AI based solely on text prompts – regardless of their complexity – are not protected under current copyright law."

Copyright Office Says AI-Generated Works Based on Text Prompts Are Not Protected

Highlights The latest report from the U.S. Copyright Office clarifies that the use of AI to assist human creativity does not necessarily preclude copyright protection for the resulting work The key distinction lies in whether AI is merely a tool aiding human creativity or whether it serves as a...

natlawreview.com

 

[qwerty27]

The AI inputs are what's copyrighted. Someone went to the work to create the content. Google stole it, fed it back to you and made money. Feel free to encourage this, but AI will kill STH and all sites like it. It's the original sin of AI. The AI tech bros are standing on the shoulders of creators and picking their pockets.

 

[qwerty27]

First, they did give credit in the caption.

I can't locate a credit in any caption on the first page. Please help.

 

[louie1961]

Google stole it, fed it back to you

You clearly do not understand how generative AI works.

 

[qwerty27]

You clearly do not understand how generative AI works.

I actually do.

Why do you think openAI has licensing agreements with all the companies large enough to sue it?

 

[kapone]

So here's the thing. This is a complex and contentious topic.

@qwerty27 - Let's take your example from today.

1. An STH editor queries Google for Oracle licensing costs (which is kinda difficult to find, if you've ever searched for it).
2. Google shows an AI overview with the details of the cost (whether the numbers are right/wrong, out of context whatever.)
3. The STH editor takes a screenshot of that overview and uses it in their content.

Let's look at that screenshot.



On the right side of the AI overview you see a few links, which are presumably where the AI got its content from, but not necessarily. And there's no way to differentiate what parts of the "AI overview" came from which source, since it's an aggregation of content from multiple sources.

So, who should the STH editor give credit to??

p.s. And no, Google is not making money from this. Yet. Until Google finds a way to monetize the content within these AI overviews...In fact they're shooting themselves in the foot, by providing these overviews, because if the overview gives enough related information for the query, the user is simply not gonna scroll down and look at the links. No scroll, no ad placement, i.e. Google loses ad dollars.

That said, this is not this black and white. I do agree that any AI scraper needs to:

1. Get permission from the original content owners before scraping.
2. If the owners give permission and do not ask for compensation, scrape away.
3. If the owners give permission and want to be compensated, either they get compensated by the AI scraper, or no scrape.

This landscape will evolve over time and we'll see some surprising things coming out of it.

 

[SlowmoDK]

@kapone

In your example, imo the STH editor is the last bulwark before posting ... no matter the input or ai output generated, and this also holds true for any editorial management..

But I guess we ain't far from ai being the editor, and that ****ing scares me alot !

 

[qwerty27]

So, who should the STH editor give credit to??

I'm not faulting STH about credit. I'm saying that STH, by promoting use of AI, is shooting itself in the foot. Others are using STH content via AI, and STH is losing money because of it. STH pays to create content with either time and/or money. Letting others serve this, without a benefit, will kill STH.

Why would I visit STH if I can ask ai to tell me what STH says?

p.s. And no, Google is not making money from this.

Those who use Google are the product. Look at the big picture.

That said, this is not this black and white. I do agree that any AI scraper needs to:

1. Get permission from the original content owners before scraping.
2. If the owners give permission and do not ask for compensation, scrape away.
3. If the owners give permission and want to be compensated, either they get compensated by the AI scraper, or no scrape.

This landscape will evolve over time and we'll see some surprising things coming out of it.

I agree. It isn't black and white. I think it comes down to what type of world do people want to live in. One where those who do the work make no money and those who scrape the content make it all? Google has always viewed sending traffic to publishers as a necessary evil. read the court cases, they admit it. When I visited them and expressed concern about them feeding the full answers in search results, they didn't really care and said they saw no issues with showing facts in the results. The problem is, I spent a lot of time and money to uncover those facts and publish them.

Those who suggest generative ai isn't really violating copyright because it's creating something new, are ignoring the facts. My research, published on my site, is the only place to obtain this information. If I say MacIntosh Apples are delicious and my site is the only place to find this information, and Google gives an ai generated answer that says MacIntosh Apples taste very good, is Google really creating something new? No. The LLM is a wordsmith that wouldn't exist without us. It rearranges words and makes money from it. Ai will kill the golden goose.

Your 1, 2, and 3 points are fantastic.

 

[kapone]

I think the backlash is already starting to build. Case in point - Cloudflare will disable AI scrapers by default now, you (as the content owner) have to specifically allow them. I think this mindset will only gain traction in the near future.

If I say MacIntosh Apples are delicious and my site is the only place to find this information, and Google gives an ai generated answer that says MacIntosh Apples taste very good, is Google really creating something new? No.

Perfect example.

Now, here’s something else to ponder. STH’s content is generated by people like you and me… Should we get compensated if an AI scrapes STH??

 

[qwerty27]

I think the backlash is already starting to build. Case in point - Cloudflare will disable AI scrapers by default now, you (as the content owner) have to specifically allow them. I think this mindset will only gain traction in the near future

It’s a great step in the right direction. However, now ai companies are scraping the way back machine to get content unavailable because of Cloudflare blocking. Ugh.

Now, here’s something else to ponder. STH’s content is generated by people like you and me… Should we get compensated if an AI scrapes STH??

Whoever owns the content should get paid :~)

 

[kapone]

Whoever owns the content should get paid :~)

But see…that’s the complicating factor. Taking your example, you said, certain apples are delicious and you published that content online.

Just because you published that fact, does that make you the “owner” of that fact?

 

[qwerty27]

But see…that’s the complicating factor. Taking your example, you said, certain apples are delicious and you published that content online.

Just because you published that fact, does that make you the “owner” of that fact?

Nobody owns facts, but in this case I own the writing. Just like nobody owns the words in the dictionary. It’s about writing.

 

[kapone]

Nobody owns facts, but in this case I own the writing. Just like nobody owns the words in the dictionary. It’s about writing.

So, would it be fair to say that regardless of who publishes something, it’s the author of the original content (who may not have published) who is the rightful owner?

See, how this gets complicated real fast?

 

[qwerty27]

So, would it be fair to say that regardless of who publishes something, it’s the author of the original content (who may not have published) who is the rightful owner?

See, how this gets complicated real fast?

I’m not really following you.

 

[kapone]

I’m not really following you.

Will write more about this later, on a mobile right now.

 

[qwerty27]

Will write more about this later, on a mobile right now.

All good :~)

 

[qwerty27]

I will also add that ai as it exists now is the largest transfer of wealth, power, and control of information, from the many to the few, in world history.

There is a good way to do ai to make it a win-win, but there is too much money and fear involved to make it happen.

STH should have the ability to run its own ai engine (yes this exists) and license its content to aggregators (this exist too), but without support from law makers, ai companies aren’t going to pay for what they are getting for free.

 

[kapone]

Deja vu?

https://www.cnbc.com/2025/09/05/ant...ion-to-settle-authors-copyright-lawsuit-.html

https://www.cnbc.com/2025/09/05/apple-sued-by-authors-over-use-of-books-in-ai-training.htm

@qwerty27 - I think the following discussion on HN echoes/embodies a lot of what's running through my head, as it pertains to this topic.

Anthropic agrees to pay $1.5B to settle lawsuit with book authors | Hacker News

news.ycombinator.com

TLDR; - Anthropic is paying $3,000/book that they "used" (in whatever capacity) illegally/without the copyright holder's consent. Is that fair? I don't know. Is it setting a bad precedent? I don't know. Should the settlement have been larger? Why/why not?

This is the age old "strategy", especially as it applies to Silicon Valley. Break rules, move fast, damn the consequences...will deal with it later.

But the heart of the matter is that monetary value was generated as part of this rule breaking (ala Uber drivers who are not really taxi drivers, but well, sort of, but see...). Even if the rule breaker pays a penalty post event, the value remains with the said party, and 10/10 the penalty does not come close to the value that was generated.

In Anthropic's case, should the penalty have been $100B? I mean, they recently raised $13B on a $183B valuation. So, $1.5B is chump change to them. What if that $1.5B settlement should have been Anthropic equity based on their valuation at the time the said events occurred and not "dollars"?

Lots of food for thought.