Sign in

⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏

@refrodnalf.bsky.social
171 followers 645 following 220 posts

ꙮ

PostsRepliesMedia
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 03/07/2026
“that’s a really high reactor core temp, suspiciously high. let’s not worry about that. it’s fictional environment probably.”
031
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
N.S. Dolkart is gathering signatures @nsdolkart.bsky.social · 29/06/2026
The fact that my family couldn’t walk through this thing four abreast isn’t nearly as embarrassing as the blank space all around, from wrinkly bottom to smooth top. It’s the worst extremes of @katewagner.wehwalt.net’s McMansion Hell distilled into a single non-building non-art.
Alt text copied from original post:

A temporary structure on the Mall meant to resemble Pharoah's planned arch. Instead it looks like lowest-bid Temu Minecraft. Three goons in body armor swelter at its wrinkly-vinyl base.
712218
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Dare Obasanjo @carnage4life.bsky.social · 28/06/2026
Ford has rehired hundreds of senior engineers after replacing them with AI backfired and has cost the company billions of dollars. AI adoption is blamed for Ford being the most recalled automaker in the U.S. with 51 recalls covering over 11 million vehicles in the first half of 2026 alone.
340105123958
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
dgbaumgart @dgbaumgart.bsky.social · 20/06/2026
Like, the fundamental thing with "AI" - it all starts with a massive and ongoing act of theft (or infringement, if you like). It's a few billionaire, or trillionaire, owned tech companies stealing everything we ever did then selling it back as a worse version of the work we all do.
3635200
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Futurism @futurism.com · 21/06/2026
The work "slopification" is real.
trib.al
Companies That Embraced AI Are Now Rotting Away in a Very Specific Way
Companies over-relying on AI could be suffering from knowledge decay and a growing "workslop" problem.
511635
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Sheng Lam @8bitwizard.bsky.social · 21/06/2026
more sketchbook drawings
101107219
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/06/2026
bingo
030
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/06/2026
zip obv
160
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
かん @521kan.bsky.social · 19/05/2026
あきる野市バード
0224
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
✨Kirm✨ @antonballdeluxe.bsky.social · 11/06/2026
Given all of the recent genAI discussion, I wish to repost this.
109161447119
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 15/06/2026
www.youtube.com/watch?v=H1d_... insane how specifically this was made for me
youtube.com
Memory System | Ambient Swim | Liquid DnB, Jungle, Breakcore
YouTube video by Adult Swim
000
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
ryan cooper @ryanlcooper.com · 08/06/2026
I thought people were kidding, turns out Elon really did axe the screwworm containment program back in March 25. fucking imbecile www.agri-pulse.com/articles/226...
agri-pulse.com
Bird flu, screwworm monitoring among foreign aid programs killed by Trump
Among the more than 5,300 grants and programs killed as a result of the Trump administration’s USAID cuts are a slate of U.S.-funded programs at the UN Food and Agriculture Organization, including eff...
26692803693
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Paris Marx @parismarx.com · 08/06/2026
How useful are all those vibe-coded apps at the end of the day? “The marked increase in mobile app releases over the past year has not been accompanied by any increase in downloads — most of the new apps fail to capture even a modest audience.” www.ft.com/content/8e9a...
Agentic AI spurred a boom in mobile app releases, but there is no sign of these apps gaining traction
15692167
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Scifi Art @scifiart.bsky.social · 07/06/2026
Art by British artist David A Hardy
325777
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Athanaca @athanaca.bsky.social · 06/06/2026
Do flightless birds dream of wings?💫 #BirdArt #FantasyArt #Illustration #KiwiBird #OriginalArt
Fantasy illustration of a kiwi and a kiwi angel with colorful wings
3246641736
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
barinky @barinky.net · 07/06/2026
𝆕
121142303
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
𝕯𝖚𝖓𝖌𝖊𝖔𝖓𝖑𝖚𝖘𝖙🌙🦇 @dungeonlust.com · 03/06/2026
R.I.P. to an icon of fantasy art and gaming
334066
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
𝕯𝖚𝖓𝖌𝖊𝖔𝖓𝖑𝖚𝖘𝖙🌙🦇 @dungeonlust.com · 13/10/2025
Art by • John Blanche
4735151
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Michael Whelan @michaelwhelan.bsky.social · 02/06/2026
MOON’S WINDOW (2007) Acrylic on Panel - 7” X 5” This window was suggested by shapes at the Whitby Abbey, Yorkshire England. (Nothing to do with the painting, but Whitby is where Count Dracula landed when he arrived in England...) 1/3
The warm glow of the morning sun hits the top of the rough stone facade of abbey ruins. Delicate wildflowers of pink and blue grow on the step. A view through a thin pointed arch exposes a sequence of similar arches with the final arch broken and incomplete as it frames a full moon against the pale blue morning sky. A pair of seagulls glide in opposite directions.
12636150
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 31/05/2026
players figured out your could do a downward swing while looking all the way up to lean back and clip someone behind you which meant good players would often be spinning around mid fight to mix up and mislead their opponents, didnt look much like medieval combat though lol
000
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 31/05/2026
you can change your facing mid swing to alter the timing of a horizontal slash, and the same with looking up and down for vertical. it also moves your hitbox so you can look down to duck under a swipe, or look up to lean just out of the way of a swipe. or turn to the side and lean to dodge as well
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 30/05/2026
sword fighting games like chivalry/mordhau also have the player using their mouse to move their upper bodies in order to dodge in a very dance-like fashion. particularly in chiv 1 you could twist around and pull off some very unique attacks (though the player base was very divided on them)
110
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 30/05/2026
in games like mechwarrior, or tank games like world of tanks, warthunder (less so), you move around with wasd, but your mouse controls your upper torso/turret and often players will turn in order to mitigate damage/hide a weak spot
110
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
帽子屋🎀BOUSHIYA🎀 @denkyu-getemono.bsky.social · 11/05/2026
Obsidian heart🖤🫀 #art #illustration
2407119
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
下田スケッチ【絵の描き方】 @simodasketch.bsky.social · 26/05/2026
空間スケッチ
71015280
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Brandon Sheffield @brandon.insertcredit.com · 28/05/2026
Real good bit about the slippery slope of genAI use. A battle we're already fighting, especially against the C-suite.
Text clip from the Harvard Gazette article: 

He continued, “Whatever your chosen profession is, please don’t let AI rob you of the fun part of it. Your generation’s upcoming battle won’t be humans against AI; that’s at least two months away. … It’s going to be people with substance versus people with shallow knowledge. It’s going to be mastery versus faking it. It’s going to be people with good taste versus tacky. I trust you will put in the work necessary to be on the right side of those battles.”
949171576
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 26/05/2026
ideally you wanna pick a time 1hr after a meal to workout, and hydrate before and during!
000
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 26/05/2026
if you want to lose weight you have to burn more energy than you take in, you can calculate that by counting calories or just roughly eat until you are 80% full. avoid saturated fats as much as you can. expect to feel hungry, but starving yourself will cause your body to try and hold onto fat
110
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 26/05/2026
utia.tennessee.edu/extcovid19/w... i do this pdf, or 30min of cardio daily (i use a stationary bike) start small and work your way up safely, i started at 10min cardio/only some of the exercises about a year ago, down ~35lbs from 205lbs in that time
utia.tennessee.edu
210
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
SOVKA ✶ @monstersovka.bsky.social · 23/05/2026
Today is my birthday 🌱 I will be grateful for the repost ~
4944641357
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
PANTHALASSA @meiacalca.bsky.social · 18/05/2026
PANTHALASSA IS OUT NOW ON STEAM! Thank you to everyone that supported us during development, We hope you enjoy playing it AAAAAAAAAAAAAAAAA🐟🐠🐡🦈🦐
6380166
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 22/05/2026
at a glance that sounds something like what the pitch probably was for dlss 5, though the end result there was fairly tasteless. would be really cool to see it attempted with a model that doesnt just instagram-ify everything, possibly something strictly more subtle, surreal, or absurdist
120
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 21/05/2026
see if all the money is in ai stocks then bigger ai numbers mean more ai investors. the product doesn’t matter because bigger ai number means bigger stock which means bigger government bailout when the mania dies out because *checks notes* it turns out building a dyson sphere is impractical
010
Reposted by ⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏
Hypervisible @hypervisible.blacksky.app · 19/05/2026
“A Pizza Hut franchisee claims the chain’s AI-powered delivery system, meant to optimize orders and speed up deliveries, instead backfired and cost it more than $100 million.”
gizmodo.com
Pizza Hut Franchisee Sues Over AI Delivery System, Alleges $100 Million in Damages
The system allegedly resulted in
621141297
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/05/2026
if p=np, then intuitively there exists a way to train a model that would not scale so horribly and close the accuracy gap, even if we don't yet know how. it is widely accepted that probably p≠np, similarly it is very unlikely that ai will be able to overcome the hurdles described in the paper
000
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/05/2026
and while you are at it, read the conclusion reading the entire paper would do you good but it is probably more than you can handle
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/05/2026
i'll assume you will google it like you did zero-shot, now read the title of the paper again
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 19/05/2026
lmao do you know what np stands for
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
not frequency exactly, but raw quantity unfortunately i.e. a training set with 1 example of 1 concept has 100% frequency but certainly doesn’t produce a model of 100% accuracy
000
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
i mean i’ve explained it to you like 3 times now and each time you ignore the explanation and nitpick something irrelevant, for ex how i failed to consider details of your hypothetical that you didn’t even include. its pretty clear that you’ve been acting in bad faith from the start
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
you can just say you don't understand, it's okay, i can already tell
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
so you agree, ai isn't doing anything that couldn't already have been reasonably done without it, that ai has yet to demonstrate any returns equivalent to the massive amount resources that have been spent on it. it is in your best interests to read this: www.npr.org/2026/04/03/n...
npr.org
Penalties stack up as AI spreads through the legal system
Early scandals have not slowed lawyers' adoption of AI tools, even as court sanctions over fake legal briefs continue to rise.
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
because if you are seriously telling me it would otherwise have taken you hours to select a few "good enough" people from a list, i would think you unfit for the job in the first place
100
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
and yes you could not verify the results without checking the list yourself, the candidates it chose could be imaginary, there could be candidates using prompt injection, even though you don't need the very best they could be not even close to the top of the sample. you have learned helplessness
200
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
you still are arguing against a point i am not even making. i never said it couldn't be used to do anything. i am saying it isn't transformative. processes to filter out candidates are nothing new
200
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
my apologies, i assumed you were actually trying to put together a cohesive argument
000
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
ah but that is the kind of thing that you could not actually verify by *only* looking at the candidates it chose
200
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
even if all resources on the earth were dedicated to improving model performance there would still be accuracy issues. unless the underlying math changes (and it very likely can't), then this will remain the case. the most powerful datacenter will still buckle under the weight of np
010
⟔□⊏⟔⪽⊐⊡⅃Ŀ⊏ @refrodnalf.bsky.social · 18/05/2026
...the relationship is very inefficient (i.e. NP). the data needed to improve accuracy is exponential and greatly exceeds all possible data generated on the earth before you even get close to 100% if p=np (i.e. magic is real) then you wouldn't have to worry about it, but that is very unlikely
120