I'm at the point that I save a copy of any sufficiently cool image I see online (for art inspo or even if it's just something that makes me happy) since I've lost pictures into the churning froth of twitter by itself after making the mistake of looking away for an hour.
*finds cool thing on any non-FOSS social media app*
Cool, I'll check this out in a minute or two.
*turning phone screen on later, app shows exactly what I was looking at, perfect to continue where we left off, but then suddenly does loading animation, goes to home page with no way to return to what I was looking at*
Yep be your own archive. Every few weeks slap that download folder into a deep storage drive. Name it with basic descriptors.
Oh shit that thing I remembers can't be found or the site went through enshittification for their IPO or government said something bad. That cool youtube video from 10 years ago has been degraded to 140p ..I thought it was in 1080p back then. Thanks gooogle!
Ok to the personal archives. Search well there's the copies. Lets upload that to whatever is [not shit]
Nah, AI is probably the greatest thing that could've happened to our ever increasing enshittification of the corporate web. All it's doing currently is just destroying algorithms that never served genuine content anyway, as neither the algorithms, nor AI, can be used for anything relevant. Google really dug their own hole on this.
I saw the writings on the wall and quit the SEO business two years back because of exactly this reason.
To be beholden on the whims of google is not a foundation to build a business around. If your entire "business" is based on collecting breadcrumbs handed to you from google compared to what they generate on what you do, your business isn't a business, it's an employment, to google, without a contract. Where they have every right to fire you at any moment when they see fit.
You're right, of course, but LLMs also are impacting the entire system of intellectual property that we built so many different systems on. I mean, I always wanted IP abolished, but it's still funny to me that the capitalists developed an automated tool for destroying their own ability to commodify artistic and intellectual labor.
Zoomers range from 11-26 years today in the year 2023
A lot of that younger half is just too young to really remember a lot of the stuff they did experience. It's like how I can't fully appreciate the pre-internet world, even though I technically grew up in it--because I was only 7 when internet connectivity got big in the US. The internet and other tech still had huge impacts in the 2000s, before this teenagers didn't have AIM messenger, or texting cellphones, etc.--and I grew up with these things being ubiquitous in my adolescence
But yea the oldest 2 years of that age range is similar to where I am
Idk, man. There was a brief period from, maybe 2002 to 2015, when Google search results were somewhat relevant. This was after the late-80s/early-90s, when web crawlers were a crap shoot at best. And the Alta-Vista / Lycos / AOL / Yahoo era where you could get high profile results reliably but everything else was 50% porn-categories on any conceivable topic.
Even at its peak, Google wasn't terribly useful outside of the first 10-20 returns. Bing was a flop, until they poached and functionally rebranded as DuckDuckGo. One of the primary appeals of sites like Reddit stemmed from the community oriented information exchange - you could go to a boutique sub on a particular topic and discover sites and content producers you'd never heard of before. Then those sites would gain prominence and start showing up in search engine queries over time. Wikipedia was, similarly, useful in large part because the information was consistent and reliable from month to month and year to year. One of the better things Wikipedia did was to lock down their bigger topics and limit who could post, if for no other reason than it curbed the "newer is better" impulse of modern web traffic engines.
But gaming search engine queries has always been a thing. Bush's guys figuring out how to get John Kerry's website to show up when you searched for "Waffles" was a thing back in 2004. Or Steve Lener rigging his Albino Blacksheep site to make "French Military Victories" redirect to a message asking if the user was looking for "French Military Defeats". Google-bombing has been around since the firm's inception. Nothing about the above is particularly new or disturbing, save in so far as its a problem these hubs of genius still can't solve.
Idk, man. There was a brief period from, maybe 2002 to 2015, when Google search results were somewhat relevant.
True, I experienced this and I'm a millennial. Plenty of zoomers did too, but they were just a lot younger I'm guessing it was only perceptible to the oldest ends of their generation group
The google results seemed to really start falling off around 2019 for me
This should for sure be intellectual property infringement... A sitemap of thousands of articles, their own SEO work, put through a plagiarism launderer? And then they publicly admitted they did so. The original site should sue.
This is literally just exactly what SEO people have been doing for the last 15 years.
There is absolutely nothing new here except for using the word "AI" instead of "article spinning". Which used to be considered blackhat practice lol.
Literally everything in this tweet thread has been standard practice among SEOs for at least 15 years. The change in the quality of the "internet" has occurred because of google getting lax on their standards and algo, not because of the SEOs. Back in the day all the spun articles would either be for niche content nobody was producing much for or it'd be page 2 stuff.
Not defending this shit. Just saying there's actually nothing notably new in this tweet thread as it's an area of marketing I actually know very well. Back in the day instead of feeding a competitor's article into an "AI" and saying "rewrite this article" we used to put it in a spinner and the spinner would rewrite by using thesaurus and heuristics to rewrite paragraphs into new ones. It'd come out a little janky but it worked for serps so meh. That used to all be considered blackhat work but over time Google got lax on it and here we are, now it's shit.
Yeah this is literally just an ad for the AI article-writing service this guy runs - and honestly I could imagine a lot of marketing shops are trying to run the same thing with just OpenAI API calls already, it looks like his prompt is just "Write an article based on this URL", which is pretty garbage considering there's more sophisticated ways at milling out bullshit like this.
Article pages that run ads. And building networks of high ranking high trust pages that can be used to link to other things for transferring trust. Ranking other stuff highly too
Sitemap is like all the links and names of links on a website. They then filled all those links with ai genned garbage. At least i think thats what it is
New levels of search enshittification poisoning results. Advertising arms race online just went from nuclear to hydrogen bombs.
Hey I'm looking for apples. They seems to be kind of scarce right now and I need to find who has the remaining ones a heck I'm wiling to pay a premium just to have one.
Orange developers / online marketers target your apple searches and spam the results with links to oranges and false leads of apples.
Google, Bing, other search engines that piggyback off them are now poisoned to death with spam and you aren't able to find the options to buy apples.
This isn't new, just easier to do with AI. Luxottica is at least one company known for polluting even tiny competing eyeglass business's search results with fake websites.
people used to spam articles, maybe with slight variation in phrasing/words, to create a ton of content easily. it's called "article spinning."
search engines learned to recognize this and punish websites that did it. it would hide/lower them on the search.
this person took all the website links from his competitors, spammed fake internet articles with those links, and now the search engines are punishing the competitor by hiding the competitor's links/website.
the "new" is that it used to be programs or wordbots things that did this shady practice, apparently AI is just spitting the shit out now and is way easier to do.
The point of the Datakrash story is Bartmoss was wrong and just turned an essential infrastructural backbone into something utterly hostile to humanity and allowed all the corporations to set up their own propietary nets they have even tighter control over. Same thing would happen irl
It's so cynical and short sighted. The byproduct of referring to everything online as "content" without caring about what that content is. Their "how to bake a pie" example only makes sense if you don't care about how to bake a pie. Short sighted because if the "content" is LLM garbage then google or w/e can easily just generate the output. Like if their example was "tensorflow optimization" there's no way the coders would be like "yeah, that's perfect. The robot will teach me." Because they understand that LLMs give wrong information, they just assume that baking a pie is unworthy of real actual instruction. Ironically, I think the coder-tech-help-blog space is actually where you can generate nonsense content and get clicks.