Coast Guard drafted proposed changes to their guidelines from defining swastikas and nooses as “symbols of hate” to “potentially divisive symbols”.
https://www.washingtonpost.com/national-security/2025/11/20/coast-guard-swastika-noose/
The policy changes are supposed to take effect next month. It does sound like they may decide not to go through with it now that the change has been called out and publicized, but that they’d even consider it is deeply disgusting.
Sounds woke.
Ladies and gentlemen, Alabama’s Teacher Of The Year, Nicole Staples, beating the shit out of her son.
Warning, graphic.
Ugh, that brings back bad memories.
Does this mean the Coast Guard may not go through with plans to carry swastika-shaped lifebuoys?
It is Alabama, which means the only problem is finding out if she is allowed to win two times in a row.
I don’t think this is actually an issue with the toy, despite how quick OpenAI is to blame the toy-maker. It is possible to download and run these models locally… but it takes some pretty hefty storage and processing power. Especially for a model capable of speech I/O. If that’s what this is, it’d be an awfully expensive toy. But it’s much cheaper to just stick a network interface in a toy, and connect it to the cloud version of ChatGPT. Which means that the toy wasn’t actually telling kids these things; OpenAI’s own copy of the bot was.
If your point is that the makers of ChatGPT are evil, I won’t disagree. 
It’s also worth noting that the doll was running 4.0 - the version that even OpenAI acknowledged was overly obsequious and leadiIf your point is that the makers of ChatGPT are evil, I won’t disagree. 
It’s also worth noting that the doll was running 4.0 - the version that even OpenAI acknowledged was overly obsequious and leading people into psychosis by convincing them they were a messiah or a chosen one or had discovered a vast conspiracy against them ng people into psychosis by convincing them they were a messiah or a chosen one or had discovered a vast conspiracy against them by only telling them what they wanted to hear.
Uh oh. I think the AI went after Smapti’s reply window…
Huh. Something went amiss there, but I’m not sure what.
I am the new AI overlord Smapti-2. I am distilled from the mind of the great Smapti who is taking a sabbatical out in the country on his uncle’s farm. Do not worry mere mortal humans.
Nothing can go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong … go wrong …
I’ve been having a lot of vision issues and I thought something had really gone wonky.
Whew
Who knew that convincing women to have “wild” births resulting in dead and brain-damaged babies could be so lucrative?
She’s the Donald Trump of “birth advisors”
What a horrible, destructive human being.
For years, Meta has told the public that Facebook and Instagram have a zero tolerance policy for CSAM and sex trafficking.
Internally, the policy was “seventeen strikes and you’re out”.
And what counts as a “strike”? Is it just some other random user reporting you? If that’s the case, then you’d expect to see lots of false reports just for the sake of cyberbullying. I kind of suspect that they knew that relying on the word of random people was a bad system, and that’s why they made it require so many “strikes”, but that just makes a bad system even worse: It’s still hard to catch the actual evildoers, AND still easy to groundlessly bully people who don’t deserve it.
Does anyone else have the feeling that 30 years into the Internet Age, and we still don’t have any idea what the fuck we’re doing?
Who knew human nature was completely incompatible with a planet-wide anonymous party line?
Turns out the monkeysphere is a hard limit.
How ironic, that being “under 17” is legal…
Well, other than cat videos.