This Week in Discourse - Moral Panic and AI Memes
People freak out that computers will draw bad things! Also - Reddit's PR is still terrible, and Elon breaks links on Twitter.
Welcome to the weekend, internet gremlins. You survived another week of online turbulence. Congrats! Relax, sit back, and enjoy the insanity the internet produced this week.
Moral panic and AI memes
Facebook recently came out with a very cool, pretty good AI sticker tool, which can make you stickers/emoji of virtually anything you want. Now, a fun fact about Facebook is that Facebook exists on the internet, and a fun fact about the internet is that the internet is filled with deranged sickos. So naturally, people are using this new fun tool to make custom emoji/stickers of child soldiers, Mickey Mouse wielding assault rifles, fetish Sonic the Hedgehog with boobs and pregnant Shrek.
Elsewhere! Bing AI Image Creator can make you some images, using the new DALL-E 3. It’s also a very good tool that can do a lot of things, and it’s also on the internet. The hot new trend is that people are having fun using Bing AI to create pictures of Kirby doing 9/11. And also SpongeBob doing 9/11! Everyone’s doing 9/11 these days, and one enterprising young sicko even combined the weird trends by having pregnant fetish Sonic do 9/11.
Now, the AI companies involved should absolutely have realized this was going to happen. Anything available to the public internet will immediately be used in the most perverted and insane ways imaginable. Rule 34 has been around for decades at this point. Of course someone was going to use AI to make pregnant fetish Sonic do 9/11. Did we not learn this lesson after Mountain Dew asked the internet for new flavor names and the top two choices were ‘Hitler Did Nothing Wrong’ and ‘Gushing Granny’?
But what’s even more interesting than ‘The internet is filled with degenerates’ is the media coverage of it. The implicit tone of a lot of the coverage is an outraged ‘How dare they’ attitude, and it’s directed more towards the tech companies than the users making these images. There’s lots of harrumph-ing and moralizing and a lot of being scandalized as though these online culture writers have never seen controversial images on the internet before. It’s not the case that every writer/article is like that, but many of them are.
Why the moralizing tone? One thing that’s happening here is that among the set of online culture writers, by far the dominant stance towards tech companies is hostility. Big Tech is something to be suspicious of, something that is powerful and antagonistic, something that must be held to account by brave reporters like us. This is perhaps a slight exaggeration, but only a slight one. And it’s contributing to the pearl clutching we see.
But on another level, this isn’t new behavior. Every new technology goes through a phase of moral panics. Video games and DnD were going to make kids violent Satanists. Encryption and the internet would enable pedophiles. Rock music would sully the chastity of the youth. You can take this all the way back to Gutenberg and the printing press, which some critics claimed would ruin the souls of monks.
New technology can be used for useful things, for silly things, and for harmful things. Everyone will panic about the harmful stuff, naturally. But if we manage to avoid succumbing to moral panics, we always figure out ways to handle the harmful edge cases without banning or regulating the new technology out of existence and everything ends up being fine. It’s happened before and it’ll happen with AI imagery as well.
Reddit continues to enrage their users
We’ve covered Reddit’s recent PR disasters stemming from their API policy changes many times in the last few months. I commented while Reddit’s user protests were ongoing:
Reddit’s handled this entire situation terribly. Every time Steve Huffman opens his mouth he inflames tensions and makes Redditors angrier. Their PR strategy here has been abysmal. The best way to handle this would have been to announce the API changes and then do and say nothing. No disastrous Steve Huffman AMA. No insulting press tour, no threats to mod teams. Radio silence. A fire only burns when it has fuel continuously fed to it, so stop giving them fuel. The online mob has an incredibly short attention span. Some new controversy will pop up in a week and everyone will move on if you let them.
During the crucial first 72 hours of the API PR crisis, Reddit repeatedly made the mistake of doing tone-deaf things that only made users angrier. And a few months later, the controversy has finally passed… so I guess it’s time for Reddit to stupidly anger their users for no good reason again.
This episode’s drama revolves around the Reddit Gold system. This was a system where users could use real money to buy Reddit coins, and then use the coins to award posts and comments with ‘Reddit Gold’ and other celebratory awards. A month ago, Reddit announced they were ending Reddit coins and awards, and that users had a few weeks to use any existing coins. This didn’t cause too much drama, although some users were disappointed as they’d purchased a lot of coins.
Now fast forward to this week - Reddit announces they are bringing back gold in a revamped format!
Now, there are some good ideas hidden in this PR disaster. The new gold/awards system allows you to buy Reddit Gold for a comment/post directly at the post, without having to go through the annoying middleman of first buying coins and then using the coins to give the post Gold. One click shopping! Makes sense, less friction.
But beyond the more convenient purchasing method, the basic description of what happened is that they took away people’s purchased Reddit coins/gold, gave them nothing for having lost those coins, and then relaunched a nearly identical system. Obviously the impacted users were angry! And to add fuel to that situation, Reddit also announced a contributor program which seems nearly identical to Twitter’s revenue sharing program. Reddit users that generate lots of upvotes and Reddit Gold donations will receive real cash proportional to the attention they get.
I suspect that Reddit thought this would be perceived positively, but it’s just another instance of the site being woefully unable to understand its own userbase. The overwhelming majority of reddit users - even power users - are not attempting to go viral on the site and rack up massive amounts of karma. They’re just fans of specific communities and Reddit happens to be the best place for that community to gather due to the site’s layout and network effects. The comments express two main ideas: skepticism that regular users will ever benefit as contributors, and the expectation that users chasing monetary rewards will degrade the quality of content on the site.
This is another step in the enshittification process that we’ve been watching so many social media sites go through. Reddit and Twitter in particular have always struggled with profitability because it’s harder to serve ads to text-heavy sites than it is for video and image based sites. So they’re seeing the enshittification hit the hardest, and it’s no surprise they’ve both had the same harebrained idea that revenue sharing with users will somehow lead to more money or better content.
A final sidenote: In a classic Reddit admin move, after the anger from the new gold/contributor system the smart move would have been to lay low and give things time to blow over. I said it months ago - when the fires are burning, stop giving them fuel! Rather than doing that, the admins waited one day and announced they would no longer allow users to opt out of ad-tracking their data. Never change, Reddit admins.
Sadly, Twitter
This was an exceptional week for stupid shit happening at Twitter, even by Elon’s new standards. Let’s do a quick run through:
There’s circumstantial evidence that Elon may have been following a white nationalist blogger’s playbook when taking over the site.
Remember all those lawsuits from fired employees? They’re starting to move through the system and Elon’s now being forced to pay their legal fees.
X/Twitter face a lawsuit from the people who claim they were already using ‘X Social media’ and had it trademarked.
Ad revenue continues to decline in what sure looks like a downward spiral
Linda Yaccarino tried to counter the declining ad narrative, noting that Visa was a ‘returning advertiser’. What was unsaid was that Visa spent ten (10) dollars. Ten total dollars.
The SEC will force Elon to testify in a potential case against him and his acquisition of Twitter
The site continues to be a buggy mess - analytics are mostly broken, and there’s bad copy floating all around the site like “the our rules”
And the big one - Elon personally insisted that the site remove headlines on links, leaving only images.
This one is the most interesting piece of news. Elon gets sued a lot, he loses a lot, the SEC is always mad at him. *Yawn*, old story. Ad revenue is down and the site is buggy and Elon has weird ambiguous ties to racist stuff, not really breaking news. But the link change degrades one of the most central features of the site - Twitter’s role as a place to find content from elsewhere on the web. The new format is just an image, and requires a text explanation of what you’re even clicking on.
Twitter’s users handled this with the tact and grace you’d expect:
The stated reasoning is that Elon personally thinks the more minimalist link images look nicer. It’s just his personal aesthetic sensibility dictating the functionality of the entire site. Going deeper, it’s likely that Elon is trying desperately to juice Twitter’s falling metrics and thinks that minimizing links is one way to do that. After all, when people click a link they’re leaving Twitter! If he makes links less useful, people will stay on the site longer!
It’s such a weird, obviously terrible move that it’s hard to provide much commentary beyond This Sucks. The man himself has also said he wants people to post content directly on Twitter, but that’s just an insane goal. Reporters are not going to stop filing stories on their websites and post 600 word reports directly on Twitter. Nobody wants to read 600 word long tweets. That’s not what Twitter is for! You can’t force people to use the site in a way they don’t want to use it.
At the core, the strategy of “Let me break my product’s functionality so that people will be forced to use it differently” will never be a good strategy. Making Twitter worse on purpose so that maybe people will change their posting style and juice your failing engagement metrics is one of the worst bankshot ideas Elon’s had for the site, and this is Elon we’re talking about. At least we can all rest easy, knowing we have at least one or two weeks before the next incredibly stupid maneuver.
just kidding one day later he announced he’d be removing likes and retweets haha everything is normal, this is fine
Links and Posts
Rebecca Jennings has an interesting piece in Vox advising you to just ignore annoying tweets. Rebecca is generally great! But as an internet culture reporter, she’s a sicko. Everyone covering the internet beat is a sicko. And so it’s no surprise that the ‘ignore bad tweets’ piece contains a 500+ word diatribe on a particular tweet that annoyed her and all the specific ways it was bad. Repeat the mantra: Nothing is stronger than The Discourse.
I say this a lot, but truly: if you want the scoop on which internet trends will be hot weeks/months before the big culture sites find them, subscribe here. This week’s example - Infinite Scroll months ago: The Rise of Girlies. Insider now: Goodbye Girlboss, Hello Snail Girl.
People are worried about Threads! Traffic is way down! Major brands have abandoned posting there! Instagram is reportedly scrambling to renew interest! When Threads debuted I wrote about how they and BlueSky were operating on different models of how social media works: Scale vs Status. As of right now, BlueSky and their high-status-limited-userbase theory keeps growing, while Threads’s instant-and-enormous-scale theory isn’t doing too hot.
Slack is, for some completely unexplainable reason, having the entire company take a week off of normal duties to reach ‘Ranger Level’ in their corporate training program.
Here’s a cool list of all the non-Western tech startups around the world competing and winning against the traditional tech giants.
A fun conspiracy about Ron DeSantis that might actually be true.
Every week we end with something positive and fun. This week’s moment of positivity to wash away the internet’s insanity is this hella cool cat.
Concerns over the labor market impact is a huge driver in the coverage of AI, and I don't think the "Spongebob did 9/11" stuff would have been covered the way it was without this lurking in the background.
This has also lead to some real oddities in how people that are normally copyright-skeptical switch if it's relating to AI.
Why did I watch that toilet thing all the way to the end. I am a fallen man.