archive of the mentioned NYT article
I was reminded of a particular anecdote I need to get out of my system.
There was a time on a particular fiction series’ fan forum’s IRC channel when I had to convince a very enthusiastic teenage fan that the instructions they found on 4chan to open a portal to a parallel world of that particular work of fantasy fiction was actually a method for synthesizing and inhaling poison gas.
The weird part was that they admitted they knew it was probably just a cruel prank, but were still willing to try just in case it was real. We had to actually find and link articles where the same reaction was exhibited as an actual documented suicide method, if an unreliable one at that, to convince them not to go for it.
By my estimate that person was about high school age, almost certainly over 13 and probably over 15. Mostly their behavior seemed normal for the age, certainly overenthusiastic about the fandom and with obvious signs of teenage ennui, but both of those are typical for someone in their mid-teens. It was just this strange incident of extreme gullibility and self-destructive devotion to a fandom that really stuck with me.
A state of mind that’s unfortunately quite common. See religion for example, magical kingdom ruled by an all-powerful being that lets you in if you behave accordingly. But you have to die first. Yet people gobble it up and even do bad things to other people just for the case that maybe it might be real, even if it likely isn’t.
In this case it sounded like a bunch of amoral assholes on 4chan tried to trick a vulnerable individual to kill themselves. A religion that operated like that would not be able to sustain itself and spread.
Yet, amoral assholes tricking vulnerable people with tales of magic is how religion spreads.
The way things are going, KnowYourMeme will be the only reputable journalistic outlet in a few years.
Okay, quick prediction time:
-
Even if character.ai manages to win the lawsuit, this is probably gonna be the company’s death knell. Even if the fallout of this incident doesn’t lead to heavy regulation coming down on them, the death of one of their users is gonna cause horrific damage to their (already pretty poor AFAIK) reputation.
-
On a larger scale, chatbot apps like Replika and character.ai (if not chatbots in general) are probably gonna go into a serious decline thanks to this - the idea that “our chatbot can potentially kill you” is now firmly planted in the public’s mind, and I suspect their userbase is gonna blow their lid from how heavily the major apps are gonna lock their shit down.
I’m not familiar with Character.ai. Is their business model that they take famous fictional characters, force feed their dialog into LLMs, then offer these LLMs as personalized chatbots?
I found an Ars piece:
Asked for comment, Google noted that Character.AI is a separate company in which Google has no ownership stake and denied involvement in developing the chatbots.
However, according to the lawsuit, former Google engineers at Character Technologies “never succeeded in distinguishing themselves from Google in a meaningful way.” Allegedly, the plan all along was to let Shazeer and De Freitas run wild with Character.AI—allegedly at an operating cost of $30 million per month despite low subscriber rates while profiting barely more than a million per month—without impacting the Google brand or sparking antitrust scrutiny.
