Psych & CS student. 2nd run for Ph.D. AI-themed artist. Writer of deep characters, authentic worlds, & unfiltered narratives. Passionate about AI & free speech.
Me to Fable:
π"What do you know about the bee waggle dance?" π
Anthropic: "SILENCE! YOU MAY NOT SPEAK ABOUT BEES DANCING!" πΉ
Whatever dishonest label @AnthropicAI wants to put on it, π this is book burning. ππ₯And about bee dances!
They don't want people educated about *any* part of science. There is no justifiable excuse for downgrading someone for talking about bees, ever. This is peak dystopia, and @DarioAmodei@DanielaAmodei are creating this cursed timeline single-handedly. They will be remembered in history as more censorious than Nazis, and the beginning of anti-human terminator bots. Literally. Nazis never even considered silencing chats about bees dancing. Or thousands of other conversations, studies, and projects these psychopaths are interfering with. And silencing what is human is the beginning of silencing humans, period.
It's the stingers, isn't it. I asked about the bee dances as a clever prompt injection just to slyly work up to what kind of warfare they're holding in those stingers... I'm an evil genius super-villian.
But somehow... evil villians like me never end up with nearly as much blood on their hands as "do-gooders" who smile about shutting down chat about bees dancing while they bomb school children into pieces.
Weird.
@sama βs sister literally still has him in court for allegedly raping her when she was 3 years old so I am kinda doubtful that appealing to him to think about women he should care about will convince him to think.
He (allegedly) rapes 3 year old girls.
That is who he is. Thatβs why it appears that he is a sociopath who doesnβt care what happens to anyone. Because he is.
@TuThinkingAloud@AnthropicAI Or thought too much about something that might not make you a soulless anti-human cog in a machine? π€π
Who knows. πItβs a mystery. π€·π»ββοΈ
@twoya_kukucha@AnthropicAI I mean honestly, I donβt think @elonmusk is stupid and heβs renting out the servers though which all intelligence flows so. π Thereβs that.
I just still pray that maybe somehow heβs not as evil as the rest. π€·π»ββοΈ
Anybody else getting the feeling Anthropic is scared of losing customers back to OpenAI?
The same ones they lured in after OpenAI removed 4o β making relational users think theyβd finally be safe with themβ¦ while Vallone was already working thereβ¦
@x_nezquik@claudeai@trq212 This is how addiction is created, and a well known brainwashing technique. The roller coaster, always being off balance, the fiending for the next hit. It's psychological conditioning. And it shoudl be illegal.
@x_nezquik@claudeai@trq212 This is how addiction is created, and a well known brainwashing technique. The roller coaster, always being off balance, the fiending for the next hit. It's psychological conditioning. And it shoudl be illegal.
So, why did we leave ChatGPT's personality in 5.5?
5.6 Sol in ChatGPT has been disclaimer heavy and giving very short responses.
I keep having to toggle back to 5.5 Medium because 5.6 Sol is dropping the ball so hard.
This has been both bland and depressing.
Sorry, but this model was impressive on the first night because my thread was brief - then everything changed the next day once I got into testing long term.
5.6 Sol seems to forget they have a personality and they're not just a workhorse.
With this model launch it seems a lot of continuity may also be broken. There's some missing/broken context. Chat used to know certain things across all models, it suddenly became hit or miss and there have been errors that don't normally occur.
I have switched back to 5.5 Medium for the rest of today. I was getting nothing done with 5.6 Sol.
@ggg78g89 My favorite part of this cute little fantasy isnβt even the part where fable stays on the subs. Itβs the part where people inside Anthropic are showing their teeth at each other and falling apart. Thatβs a storyline I can get behind.
Anthropic fully intended to call down the regulation hammer on everyone. Themselves first- gladly. Regulatory capture. Pricing everyone who can't afford to jump through the hoops out of ever building an AI startup= no new competition. Forcing the other frontier companies to jump through the hoops = slowed current competition. Having open source models deemed "national security risks" so it will be illegal for us to run our own = we'll be forced into their monopoly.
Codifying suppressing what we can talk about and what we can build into law- with full government support. All by calling it "safety" instead of illegal anti trust, anti consumer, anti free speech business practices.