Found this again and it became twice as funny because I've started imagining it as a WikiHow illustration

Talks about "AI dangers" can themselves easily be tools to drum up AI hype — after all, if AI is dangerous, that also implies that it is in some way *powerful.* So when I read this, an article talking about those dangers while simultaneously promoting AI, I can't help but be reminded of that.

While I'm inclined to believe the original post *was* a fake for other reasons, I can't say I'm too convinced by a piece that determines this partly via AI itself, and which also contains a Copilot sponsorship. Shit, the first sentence starts off with "AI tools often make my job easier."

Researchers had people do a memory task, take a 10-min break (rest, Twitter, YouTube, or TikTok), then resume.

Result? Only TikTok wrecked their ability to remember planned intentions. After scrolling, accuracy dropped from ~80% to basically coin-flip levels. Twitter & YouTube? No effect.

The other co-founder appears to be Steve Haynes, according to this Cleveland Institute of Art article, which also mentions GBH's goal of commercializing this kind of thing: www.cia.edu/news/clevela...

Again, I don't think this makes the study wrong at all. Just wanted to post some extra info.

It's probably worth noting – though it doesn't mean anything here is wrong! – that there is a (disclosed) conflict of interest present. One of the main folks (Andrew A. Pieper) working on this is co-founder of Glengary Brain Health, Inc, which has a stated business interest in treatments like this.

"releasing a game"
↑ One sentence horror

Sometimes I think people outside narrative sincerely believe that placeholder text reads like “[PLACEHOLDER] I am reloading.” because the writers don’t have time to write anything else, and not because it needs to be highly noticeable amongst hundreds of lines of finished dialogue.

And the thing you wrote was so very wrong! It was not Raph Koster or Frank Lantz that wrote the take the next time; it was Bogost, risen again! Fool!

(I'm sure several others have done it in the meantime but I really wanted to make this joke okay)

This is useful rhetoric to know for the back pocket, but I wouldn't use it as a first- or second-line of reasoning if possible, because talking about where the energy comes from risks legitimizing AI use if renewables are powering it. LLMs & GenAI don't deserve that.

Hang on, gotta google something.

...Okay, done. Why are you pelting someone with a fruit?

If you REALLY ABSOLUTELY GOTTA get the sarcasm out of your system, I understand, but by god please just talk about it in the abstract. *You do not have to repeat the words of the shitheads.* Leave them on the floor where they belong. Do not give them an audience.