quarantine 'em

4. the algorithm is startlingly good at imitating my writing style for updates at first but it wanders way off topic and starts repeating itself because, you know, it's a bot. like how predictive text challenges always get locked in a loop

(going to stop numbering these) so by the time of present-day updates it's going in psychotic circles talking for thousands of words about "metal" and "low ceilings" and it generally seems to be quite unhappy for a bot

Other current theory is that they kept this going as a semi-actual social network which they could experiment on

which is actually quite cool because Facebook had that problem where they were doing ethically dubious experiments on their users and, Google People, the users are all screaming mad bots, problem "solved"...?

of course it could be that classic "AI" thing where they're actually just woefully underpaid real people writing miserable fake posts in trash conditions

whatever's going on this latest auto generated photo of "me" looks pretty angry and like legitimately sick

"turn on your monitor" yuk yuk but seriously it's like they used a "radiation sickness" filter. it's actually legit upsetting to look at. no thank you

this rings a bell but it could have been Google Plushttps://twitter.com/zarawesome/status/1170481130933575685 …