alice ♡ ∩_∩

#cutiesaturday

You talked me into it lol

Some expressions 🐐

screenshot of patreon's payouts page

reminder when you're doing your taxes, the number patreon is putting on your 1099 is the gross and they DO NOT exclude all the money they took from you in fees. you need to deduct that yourself as business expenses. an easy ("easy") way to do that is to download your earnings by month

Fran from Final Fantasy 12, in her Woodweller's Garb from Opera Omnia

Nav Toor @heynavtoor 🚨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?

⚠️Told you so moment:

OpenAI published a paper proving that ChatGPT will always make things up [...] Always.

They proved it with math.

[...] This isn't a bug they're working on. It's baked into how these systems work at a fundamental level.

More: x.com/heynavtoor/s...

Jack Nicholson proves age hasnt dulled his sense of humor by throwing a stun grenade into a crowd of his fans

(Source: tinyurl.com/5bkf7bhh)

Everybody got their Tony

Sketch of Lash from Deadlock doing his new chest bump parry animation, only he's drawn to have a comically large chest.

Lash Big Naturals 🎨

#deadlock

[This user will not survive the boss battle.]

i know i should be strategic about ammo usage in re2 but i looooove emptying all my weapons on a hall way

im not gonna say anything but i think a company requesting a DMCA takedown of a slime tutorial and then accidentally attaching the entire fucking raw source file in the same email is probably the funniest case of lost media becoming found i've heard in ages

Goat is pissed off

Drawing of Exdeath from Final Fantasy 5

Drew my favorite evil tree Exdeath <3 #finalfantasy

I get back to drawing Persona girls now #art #persona

ive had this pair since 2019 lol
it's so Yellow

i need new glasses

japan will never have this

ok

NetEase definitely sucks but this game has also been in development for a while and with the extra budget needed, it definitely feels like there has been some sort of mismanagement. I really hope the game sees the light of day though, it’d be unfortunate if the whole thing went up in smoke.