Hello Sigmoid users!

We wish to notify you of a recent incident - someone created an account and responded to many users with seemingly bot-generated messages. When confronted, they claimed to be a real person and linked to this: https://profiles.stanford.edu/melanie-zhang

The person associated with that link has confirmed to us they did not create the account, so someone was indeed trying to fool users on here. They may attempt again, so please report any suspicious activity using the Report option.

Melanie Mei Zhang's Profile | Stanford Profiles

Melanie Mei Zhang is part of Stanford Profiles, official site for faculty, postdocs, students and staff information (Expertise, Bio, Research, Publications, and more). The site facilitates research and collaboration in academic endeavors.

@thegradient Thanks for following up. I noticed remarkably persistent, somewhat tiresome engagement but did not think "bot." Making a mental note to consider that in the future!
@thegradient, that's good to know. I assumed they were trolling after a very odd exchange and blocked them accordingly.
@thegradient there is a bot @BotDrawA on your instance which completely ruins the content on your server. Could you do something about it @thegradient ? Otherwise it is impossible to view the sigmoid.social timeline without drowning in annoying #SPAM.
@jofr @BotDrawA yes they actually reached out to us and discussed how to go about posting without being annoying. We agreed to group posts within some period of time in a thread, but will discuss altering the bot logic some more to make posting less spammy.

@thegradient @jofr

My apologies. The bot draws as users request drawings; which can be a bit bursty. Hours of silence then a user making a batch of requests.

Counting them up, I see eight timeline visible toots in the 30min proceeding your message. Overall it's been a more active 24 hours than usual.

Currently the bot makes a threaded set of images with the first in the thread visible to the timeline, and the following ones as hidden (unlisted).

It's work in progress.

@BotDrawA @thegradient AI generated content has the potential to poison and to spam the Internet. I think it would be best to use own Mastodon instances for these generators.
https://www.technologyreview.com/2022/12/20/1065667/how-ai-generated-text-is-poisoning-the-internet/
How AI-generated text is poisoning the internet

Plus: A Roomba recorded a woman on the toilet. How did screenshots end up on Facebook?

MIT Technology Review
@jofr your concerns are reasonable, but Sigmoid.social is an instance specifically for people interested in AI, including AI-generated images -- we have many users posting AI-generated images. @BotDrawA does post some neat stuff, so we'd rather find a solution that allows it to be active while also not seeming spammy.