- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
The linked article does his math wrong - the ad revenue per page view is $0.00258, so the profit per page is actually $0.00207.
This will radically accelerate the problem of search results being full of AI-generated spam.
RIP internet
It was already a massive issue pre-LLM. This shit sucks so much lmao.
Were going back to webrings and webmaster babeeee
Unironically this. I’m setting up my website this weekend that’s going to be all static html. I’d join a hexbear webring.
i am so ready
i feel like there’s a treadmill here that they’re starting us all off upon. as search results rapidly become full of ai-generated spam the odds of a pageview rapidly go down. so the whole thing ends up being a bit like bitcoin mining, trying to find that exact right combination of seo blather to be the page that’s selected by the search algorithm for the precise prompt entered by one of the last goobers gullible enough to still be searching for useful information and then also somehow convinces that goober to click on the page in order to be inundated with ads and hallucinatory garbage. ad revenue per page view will fluctuate wildly based on speculation about market sentiment about new and innovative optimization schemes, reaching stellar heights of value even as it kills the internet entirely and everyone drops their phones and laptops and starts the butlerian jihad against our ai oppressors irl just like in that movie!
good thing the only articles I read on the internet are declassified ClA documents
AI generated CIA documents when?
the CIA has brain worms so assume they have jumped on the hype.
How long does it take to FOI or whatever it is over there?
I look forward to dumps of “As an ai language model I cannot provide information on how to trick the government into funding dolphin sexperiments.”
I listened to an episode of Darknet Diaries that said law enforcement and intelligence do use ‘AI’ for scanning and summarising, I think it was this ep: https://darknetdiaries.com/episode/146/
Surely this is just overusing the term “AI” when its more appropriate to use machine-learning here. The image recognition technology to read text has been trained on people doing captchas. All those recaptchas you did where it gave you 2? One was you giving them data on something their system couldn’t recognise and the other was them actually testing you. This became the technology to read handwritten and scanned text.
Calling it AI is a stretch in my opinion. There’s no “intelligence”. We’ve just reached the point where they apply it to anything and everything if machine learning is involved in any way.
Yep exactly, that’s why I put it in quotes, because iirc in the episode itself they still call it ‘AI’ and I wanted to convey that yes, they are using that technology.
Capitalism ruins everything it touches
CREAM
This is ridiculous. It’s over. It’s so over. It’s so fucking over.
Infinite money glitch???
rest in piss to the internet, it was a shitty idea to begin with
Death to America
Only while OpenAI subsidizes GPT text generation for free. How long until they realize setting money on fire is not a profitable strategy?
If outlets had to actually run their own models at cost, it could never even compare to the revenue of one page view.