Hold of the World @holdoftheworld - Tumblr Blog

yansurnummu

@chameleonspell my brain would not let me rest until I did this

a snapshot of years later, maybe. I like to hope that Iriel's dad does adore Julan :')

the-earlof-pudding

I am a simple creature, I see Julan, I reblog Julan.

kosmogrl

let this be a reminder that even if it takes time your dreams can come true

rabbitindisguise

an astronaut foretold to go to the moon by a fortune cookie can be a kind of magical girl, I think

butchfriend

Everyone Is Cheating Their Way Through College

cognitiveinequality

sturionic

For anyone unfamiliar with Paulo Freire's Pedagogy of the Oppressed, the entire thesis is that traditional educational models promote oppression by removing students' agency in their own learning. Freire argues that currently education functions as a "banking model" - teachers are the holders of knowledge, and students are empty vessels, waiting to have that knowledge put into their heads like a piggy bank. This reinforces a passive attitude towards information, not seeking and understanding it on your own terms, but waiting for a "banker" to deposit it into your head.

Instead, Freire proposes that teachers and students act as co-creators of knowledge, where students become active participants in their own learning through questions and dialogue. Teachers are also open to changing their understanding of topics in the process of critical dialogue - the goal is not "student learns Fact A and memorizes it as presented," but instead the goal is the knowledge itself, discovered collaboratively by teacher and student, who are acting with empathy and respect towards each other. This also starts the process of the oppressed being able and empowered to question structures of power, take agency, and actively participate in the transformation of society.

So, the irony of writing an AI essay on critical pedagogy is actually insane; because it's essentially the extrapolated endpoint of Freire's arguments that our current educational system creates passive receptacles who not only can't think critically in an educational context, but also become the perfect citizens for a world that doesn't want us questioning structures of power, to view those in power as we viewed our teachers - deliverers of indisputable facts that must be memorized and regurgitated because they command it, and not co-creators of true understanding.

heedra

really losing my patience for any 'feminist' statement to the tune of 'we need feminism because women fill a fundamentally different and necessary role than men and will be better at doing x y or z'. like actually i think we need feminism because it is an unbearable death of the spirit by inches to exist in a world where you are not seen as a fully realized human being because of a single cultural determination, and because a world that enshrines such things creates systems that are fundamentally sick to the core

woman-becomer

tumblr users love reading. you literally stopped for this post just because it has words in it

interpolationz

this is one of my favorite bits about tumblr

the users seem to actually prefer text posts to anything else, and treat it as a chore to play a video especially with sound

lazzarella

grendel-menz

mommy, daddy, baby

dracula1992soundtrack

being online is so scary aren't you guys worried about the world wide spider

redfagdiver

It took 36 years for someone to make this joke and by god it was worth the wait

klfg

[source]

klfg

🚨BREAKING: OpenAI published a paper proving that ChatGPT will always make things up.

Not sometimes. Not until the next update. Always. They proved it with math.

Even with perfect training data and unlimited computing power, AI models will still confidently tell you things that are completely false. This isn't a bug they're working on. It's baked into how these systems work at a fundamental level.

And their own numbers are brutal. OpenAI's o1 reasoning model hallucinates 16% of the time. Their newer o3 model? 33%. Their newest o4-mini? 48%. Nearly half of what their most recent model tells you could be fabricated. The "smarter" models are actually getting worse at telling the truth.

Here's why it can't be fixed. Language models work by predicting the next word based on probability. When they hit something uncertain, they don't pause. They don't flag it. They guess. And they guess with complete confidence, because that's exactly what they were trained to do.

The researchers looked at the 10 biggest AI benchmarks used to measure how good these models are. 9 out of 10 give the same score for saying "I don't know" as for giving a completely wrong answer: zero points. The entire testing system literally punishes honesty and rewards guessing.

So the AI learned the optimal strategy: always guess. Never admit uncertainty. Sound confident even when you're making it up.

OpenAI's proposed fix? Have ChatGPT say "I don't know" when it's unsure. Their own math shows this would mean roughly 30% of your questions get no answer. Imagine asking ChatGPT something three times out of ten and getting "I'm not confident enough to respond." Users would leave overnight. So the fix exists, but it would kill the product.

This isn't just OpenAI's problem. DeepMind and Tsinghua University independently reached the same conclusion. Three of the world's top AI labs, working separately, all agree: this is permanent.

Every time ChatGPT gives you an answer, ask yourself: is this real, or is it just a confident guess?

sanctus-ingenium

Urtica

nemfrog

The serpent consuming its tail, crowned with flame. Frontispiece by Melchior Lechter for Stefan George's, Der siebente Ring. The Seventh Ring. Berlin, 1907.

Internet Archive

pragnificent

This is Money Snake. She only appears every 312 years.

If you reblog her picture within the next twenty-five seconds you will have good luck and fortune for the rest of your life.

thingsamylikes

I reblogged her late last year and my 2024 has been very satisfying work-wise and (secure enough to not stress out) money-wise so far. Money Snake is wise and good.