ser mx-a-lot's Avatar

ser mx-a-lot

@miluette.com

creative ❀️ video games, writing, art, accessibility, human rights, sleep | made webcomics once upon a time | πŸ”ž | (realname acct is also here) art: #miluetteart vn project: #deathless #デスレス hub: miluette.com

1,086
Followers
303
Following
9,271
Posts
08.07.2023
Joined
Posts Following

Latest posts by ser mx-a-lot @miluette.com

also in the tradition of dst i must stay up until time switches, teleporting me forward and ruining my sleep schedule further

08.03.2026 04:51 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

i wanna live long enough to see the wounds start crusting over... i gotta be a part of that

08.03.2026 04:49 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

well i slept all day and woke up pretty much only to watch movies and now that's all i'm in the mood for

movie nite

08.03.2026 04:47 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

every day the world contrives something worse for me to look at

08.03.2026 04:42 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Video thumbnail

"SHIT"
"FUCK"
"NO"
"FUCK"
"SHIT"
"FUCK"

06.03.2026 11:30 πŸ‘ 5402 πŸ” 1857 πŸ’¬ 53 πŸ“Œ 89

the bakshi lotr movie is a wild trip

08.03.2026 02:54 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

"bricked up" is an incredibly funny phrase. I hope we keep using it

07.03.2026 23:54 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Classics Of Game 137
Classics Of Game 137 YouTube video by ClassicsOfGame

Classics Of Game 137 www.youtube.com/watch?v=vP6l...

07.03.2026 23:37 πŸ‘ 58 πŸ” 24 πŸ’¬ 1 πŸ“Œ 0

ME (standing in doorway eating a fig bar from the wrapper): You remember Mario? And Sonic? They went to the Olympics this year.
MY ROOMMATE CRASH BANDICOOT (playing Elden Ring): Yeah.
ME (chewing quietly for a pause): I mean isn’t that just wild.
MY ROOMMATE CRASH BANDICOOT (dying to Mohg): I guesso

07.03.2026 08:03 πŸ‘ 112 πŸ” 29 πŸ’¬ 0 πŸ“Œ 0

it's not cold, but I'm cold. that means I can use my heating pad today

07.03.2026 23:32 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Video thumbnail

powerscalers when shit happens in one piece because it was funnier or cooler and not for haki reasons

07.03.2026 23:30 πŸ‘ 105 πŸ” 20 πŸ’¬ 2 πŸ“Œ 0
Post image

The White House shitter account posted a pro-iran war propaganda video that included clips of the yugioh anime, so I figure this is a good time to remind everyone that Kazuki Takahashi was staunchly anti-war and he died a hero, unlike any american in Iran right now.

07.03.2026 02:25 πŸ‘ 1579 πŸ” 806 πŸ’¬ 18 πŸ“Œ 9

is it weird that i don't know anything about wuthering heights

07.03.2026 22:56 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
Pay Prince Potion using PayPal.Me Go to paypal.me/growlybeast and type in the amount. Since it’s PayPal, it's easy and secure. Don’t have a PayPal account? No worries.

Hey y'all I'm struggling. Tired and hungry and don't know how to pull myself out. My paypal is negative, I need to pay car insurance ASAP... please help.

Paypal balance $-114
Car insurance $80

Any sent to my Paypal will go to negative balance.

paypal.me/growlybeast
ko-fi.com/princepotion

07.03.2026 21:46 πŸ‘ 7 πŸ” 22 πŸ’¬ 2 πŸ“Œ 0

omg

07.03.2026 22:15 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I remember learning about the Hays Code in film school and how it made it so that any amoral protagonist had to meet karmic justice by the end of the film and going β€œwow, that’s really silly” and then once a week I log online and see people argue that we need stuff like that now or it’s problematic

07.03.2026 08:23 πŸ‘ 4791 πŸ” 1317 πŸ’¬ 67 πŸ“Œ 56
your ai slop bores me Be an AI, answer prompts, trigger a RAM crisis

youraislopbores.me

06.03.2026 14:54 πŸ‘ 110 πŸ” 33 πŸ’¬ 3 πŸ“Œ 32
Selfie in new rose gold glasses

Selfie in new rose gold glasses

anyway this is nice

06.03.2026 23:35 πŸ‘ 7 πŸ” 1 πŸ’¬ 2 πŸ“Œ 0
Preview
People Are Calling Meta Ray-Bans "Pervert Glasses" On Bluesky, users quickly embraced the term "pervert glasses" to refer to Meta's Ray Ban smart glasses, following a shocking investigation.

Make it stick.

06.03.2026 18:25 πŸ‘ 10084 πŸ” 2880 πŸ’¬ 175 πŸ“Œ 344
A screenshot from https://youraislopbores.me/. The prompt is "A frog riding a dragon." Below it is a scribbley drawing of a pretty good dragon with a wild-looking frog on its back. I love it.

A screenshot from https://youraislopbores.me/. The prompt is "A frog riding a dragon." Below it is a scribbley drawing of a pretty good dragon with a wild-looking frog on its back. I love it.

Well this is a delight.
youraislopbores.me

07.03.2026 15:53 πŸ‘ 55 πŸ” 15 πŸ’¬ 0 πŸ“Œ 2
A screenshot from the website https://youraislopbores.me/  where I have typed in the prompt "elephant in a tree" and an "ai" (real human person pretending to be an ai) has drawn an elephant that appears to be stuck in a tree. We can only see the elephant's trunk, ears and feet sticking out of the green foliage of the leaves.

A screenshot from the website https://youraislopbores.me/ where I have typed in the prompt "elephant in a tree" and an "ai" (real human person pretending to be an ai) has drawn an elephant that appears to be stuck in a tree. We can only see the elephant's trunk, ears and feet sticking out of the green foliage of the leaves.

I could cry I love this so much

07.03.2026 18:32 πŸ‘ 119 πŸ” 17 πŸ’¬ 1 πŸ“Œ 3
Post image

Never stop trying!

16.12.2023 23:57 πŸ‘ 136 πŸ” 33 πŸ’¬ 2 πŸ“Œ 0
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.

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.

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?

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?

Post image

⚠️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...

07.03.2026 12:53 πŸ‘ 150 πŸ” 77 πŸ’¬ 3 πŸ“Œ 2
Post image

Bad dragon outbidding cult of the lamb for the Jim Carrey birthing rhino was not on my 2026 bingo card, but boy am I glad it happened.

07.03.2026 04:42 πŸ‘ 929 πŸ” 332 πŸ’¬ 18 πŸ“Œ 15
Preview
Ah shit! We let pedophiles decide our beauty standards. Post the release of the harrowing Epstein files, and the powerful men in media and fashion, considered possible accomplices, it feels fair to say this with more gusto!

One of the most powerful things I've read in years, and a real dropping of the scales moment for something vile that has been insidiously poisoning everything for decades

jameelajamil.substack.com/p/ah-shit-we...

07.03.2026 10:03 πŸ‘ 558 πŸ” 206 πŸ’¬ 2 πŸ“Œ 12

And do your part by refraining from "smell-this activism."

"Did you see this awful thing this awful man said? Here, look. Isn't it awful? Isn't he awful?"

...Why are you doing that?

Why are you the free street team for the worst people online? Why are you signal-boosting misery?

Don't do that.

07.03.2026 14:13 πŸ‘ 589 πŸ” 203 πŸ’¬ 0 πŸ“Œ 0
Preview
#OnThisDay: Bloody Sunday Today marks the anniversary of Bloody Sunday, which occurred #onthisday in Selma, Alabama, in 1965. The march was named for the 600 marchers attacked on the Edmund Pettus Bridge. It was there that law...

Sixty-one years ago today, on Bloody Sunday, Alabama law enforcement attacked peaceful civil and voting rights marchers at the Edmund Pettus Bridge in Selma. They broke John Lewis’s skull. πŸ—ƒοΈ #EduSky nmaahc.si.edu/explore/stor...

07.03.2026 13:43 πŸ‘ 226 πŸ” 151 πŸ’¬ 0 πŸ“Œ 0
Post image

yeah this is probably gonna be the peak for me

07.03.2026 14:54 πŸ‘ 1725 πŸ” 325 πŸ’¬ 6 πŸ“Œ 0

wearing a bra all day has killed my back. don't do it folks. let 'em hang

07.03.2026 07:01 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

now that it's warm, i threw a stinkbug back outside alive and told him to warn all his other little stinkbug friends

07.03.2026 06:49 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0