Friday, March 15, 2024

Ragebait: Why Everything You See On the Internet is AI-Generated Fakes

People click on things that enrage them, which ensures more enraging content.

When we lived near Ithaca, NY, we saw a lot of green Subarus with bumper stickers saying (among a lot of other things), "If you're not outraged, you're not paying attention!"  I talked about this before and came to the conclusion that if you are outraged all the time, you are easy pickin's for our commerce culture.  While you get outraged over things, you don't actually change them.  Meanwhile, they pick your pocket.  

It is like people complaining about inflation, as they idle their SUV in line at the McDonald's drive-thru.  All the facebook postings in the world are not going to change anything until you stop consuming.  And according to some articles (thanks, reader!), this in fact, may finally be happening, as people run out of money and credit.

But I digress.

A lot of the content of the Internet is AI-generated and you may not realize it.  While people debate about the possible impacts of AI on the future, the future quietly arrived (yes, it took me by surprise, too).  I illustrated before how a simple ChatGTP prompt can create a pretty realistic-sounding blog posting (to the uninitiated, anyway) or a very realistic Reddit posting.

And rage-baiting is the name of the game.  AITA (Am I The Asshole?) stories about cheating husbands, in-laws ruining weddings, ungrateful children, or narcissistic parents are all designed to get you upset and get you to click.  And when you click, your blood pressure goes up and you become anxious.  Or maybe you click just to see how stupid the story is. Either way, they get clicks, eyeballs, and "engagement."  Reddit is about to "drop" their IPO - no doubt the content will go 100% AI after that.

AI-created rage bait is more slowly spreading to YouTube.  AI videos are still somewhat primitive and easy to spot. No problem, though, as there are a host of "influencers" out there, willing to make fake videos to enrage you.  Ever wonder why everyone hates "influencers"?  It is by design.  It is the old Howard Stern "You'll never guess what he said today!" outrage.  They don't listen because they like him, but because they want to be outraged.

It ain't hard to find an older relative to play the part of "Karen" and go off on camera.  In almost every case, you have to wonder why the cameraman is filming in the first place - watching paint dry, perhaps?  Or even if "Karen" is real, they never show what happened before the video started.  Many of these idiots will actually set out to annoy someone to get a reaction and then film it. We all suffer as a result.  It is "Candid Camera" gone off the rails.

People learn intuitively, what works and what doesn't.  When I monetized my blog many years ago (for a whopping $2000 in one year) I learned that anything I wrote about Hillary or Trump was cash in the bank.  Major American news outlets have learned the same thing.  Controversy sells, whether it is people curious as to what the controversy is all about, or people rooting for one side or another.  It doesn't matter either way - a click is a click. So, online "influencers" learn - explicitly or intuitively - that being an asshat to everyone sells clicks and puts money in their pocket.  An online "personality" who plays nice, sinks to the bottom of the Internet septic tank.  Go controversial or go home.

It is akin to how quickly the poor figure out how to benefit from welfare programs.  People who barely passed the 8th grade can recite, from memory, United States Code sections dealing with public benefits programs - and tell you all the qualifying criteria and what forms to fill out, where to file them and when to expect your first check or benefits card.  When it comes to survival, people learn fast.

Similarly, the homeless quickly learn what works and what doesn't, in terms of begging strategies, free handouts, good places to sleep without being disturbed, where to get drugs, and so on and so forth.  It amazes me that I can cross a continent and see a homeless guy on each coast with the identically worded sign.  A Madison Avenue marketer armed with a dozen focus groups could not do better.  People do what they need to do to survive and quickly figure out and share these tips.

With the Internet, however, it seems this intuitive learning process has discovered a dark side of humanity - that what we will likely click on is stuff that makes us angry, depressed, and anxious.  Few, if any, click on the "heartwarming story" or if they do, it is only to leave a snarky comment.  Speaking of which, the comments sections on most of these sites are the worst part of it.  People leave the most mean-spirited comments, safe behind their wall of Internet anonymity  They do not profit from this, other than in terms of attention-getting.  And like others listed above, they quickly learn what gets them the most attention - saying the most horrible things.

Again, with AI and foreign influencers (which includes our own CIA, which was recently "outed" as running an influence campaign in China - act shocked, I know) the entire process is being bootstrapped literally at the speed of light.  Programming a "bot" to prompt ChatGPT (or other AI site) to create rage-bait postings isn't hard to do, and you could literally generate thousands of such postings per hour, drowning out any enlightening comment or indeed, any human-generated content.

There is precedent for this, as I noted before.  My first experience with the Internet back in the 1980s, was with usenet discussion groups online, which were reached using an ASCII-text terminal interface.  It was mostly computer nerds exchanging opinions about which Star Trek character was their favorite or whatever other drivel.  Advertisements were frowned upon and shouted down.  But over time, they overwhelmed many newsgroups, which essentially shut down as real users fled to private websites and the newsgroup left behind became 100% SPAM.

Moderation (having a human "moderator" monitor all postings) is one option, but is problematic in its own right.  Since most of these sites make so little money, there is no way to pay a moderator.  As a result, most are volunteers who get tired of the flame wars rather quickly.  Keeping up with a firehose of SPAM is nearly impossible, particularly when it is aimed at your group 24 hours a day.  Some moderators get a little power-hungry too, and censor or ban users who they simply don't like or have opinions contrary to their own.  It is a real mess.

The use of real names (as opposed to made-up user names) was thought to be a panacea and Facebook went this route.  Sadly, it turns out that a lot of people have no problem putting their real name to odious or stupid opinions, such as embracing Nazism or promoting flat-earth or anti-vaxxer beliefs.  Checkmate, fact-checkers!  Of course, in some cases this has resulted in people being "outed" as bigots or racists and thus losing jobs or friends.  And this lead to the fictional "cancel culture" nonsense - predicated on the notion that no matter how odious you are, you are entitled to a job, friends, or even a spouse (the latter according to the incel subgroup).  Everyone has rights except you and I, it seems.

Again, controversy sells, and oftentimes, what starts out as ragebait or even a joke, ends up being taken seriously by some folks.  The whole "incel" thing, for example, I think started as a joke, but people ended up self-identifying with it.  Flat-earth, I think, is the same deal - perhaps "chemtrails" as well.  The whole "birds aren't real" thing has yet to be taken seriously, but I think that is only because the creators of it have gone out of their way to make it clear they are mocking conspiracy theories with it.  But crypto-currency?  Probably a joke that went too far.  Dogecoin literally started as a joke but now is being hyped by the "richest man in the world."

And grifting falls right in with this.  By hyping a "coin" or stock or whatever, online, you can be sure that some idiots will buy it, which in turn will drive up the price which in turn will make you rich, if you bought that stock for cheap (or an option) before you hyped it.  This sort of shit used to be illegal, but on the Internet, it is hard to track down anonymous accounts, particularly those from overseas.

The image above was made using an AI program with the prompt, "Show a picture of an African boy with a car he made out of plastic bottles" or something similar.  It was posted (and re-posted and shared and forwarded endlessly) along with a heartwarming ChatGPT "story" behind it, lauding the resourcefulness and creativity of this poor African child who made something from nothing.  Of course, it never happened.  It is all fake.

People clicked on the story to leave outraged comments about AI-generated content, which in turn, pushed up the click-rates and site engagement.  Mission accomplished.  And here I am, further raising awareness of it.  You can't win at this game - it is rigged.  The only real solution is to spend less time online, less time on social media, less time clicking on click-bait links.  It is harder to do that than it is to quit an opioid addiction, I think.  And hundreds of millions of Americas are hooked - billions, worldwide.

Doom-scrolling the smart phone has replaced the cigarette-break as the number one destroyer of productivity, worldwide.  A smoker might take five or ten minutes on a smoke break.  An Internet addict can spend hours - a whole workday, in fact - falling down the rabbit-hole.

And try to take away their phones, as we once did in schools.  People will howl and likely resort to violence before they give up their electronic drug.

Not before, of course, they make a rage-bait posting about it!