microbians @microbians

Gabriel Suchowolski ☻ Creative Design Senior Director — Co-Founder @domestika ☺︎ https://t.co/d7XltrVMV3 ☺︎ @robotscrytoo @bimbadabum microbians.com Metaverse Joined September 2009

Tweets

38K
Followers

7K
Following

2K
Likes

50K

screencaps @screenscaps

13 hours ago

backrooms (2026) x exit 8 (2025)

96 4K 44K 1.3M 5K

View Details

You have noticed it. ChatGPT feels dumber than it used to. Your prompts that worked six months ago produce worse results now. The writing sounds flatter. The ideas sound safer. The internet itself feels like it is shrinking. Every article reads the same. Every email sounds the same. Every answer sounds like it was written by the same voice. You thought it was you. It is not you. Researchers at Oxford and Cambridge published a paper in Nature proving what is happening. They call it Model Collapse. Here is the mechanism in one sentence. AI trained on AI-generated data gets dumber every generation until it forgets what real human data looked like. The internet is filling with AI-generated content. Blog posts. Articles. Reviews. Comments. Social media. AI companies scrape the internet to train the next generation of models. Which means the next generation of AI is being trained on the output of the current generation. Each cycle loses information. Not randomly. It loses the rarest, most unusual, most creative parts first. The researchers call these the "tails of the distribution." The weird ideas. The unexpected perspectives. The things that made the internet feel human. Those disappear first. What remains is the average. The safe. The expected. The bland. Then the next generation trains on that. And loses more. And the next generation trains on that. And loses more. The researchers proved this is not a slow decline. Major degradation happens within just a few iterations. Even when some of the original human data is preserved. They tested it on large language models. On image generators. On statistical models. The pattern was the same every time. The output converges toward a narrow, flattened version of reality that looks nothing like the original data. The lead researcher put it plainly. "Large language models are like fire. A useful tool. But one that pollutes the environment." The pollution is invisible. You cannot see which sentence on the internet was written by a human and which was written by AI. Neither can the AI that is about to train on it. And once the tails are gone, they do not come back. The damage is irreversible. This is not a prediction anymore. It is a diagnosis. The internet you grew up on was built by humans writing things no algorithm would have written. Strange, personal, imperfect, alive. That internet is being diluted. One generation of AI at a time. And the models trained on what remains are learning a smaller and smaller version of the world. Model Collapse is not a technical problem. It is a cultural one. The thing that made the internet worth reading is the thing that disappears first.

431 3K 6K 328K 3K

View Details

まだ面白い @madaomoshiroi

a day ago

人類史上初めて宇宙探査機が太陽コロナに突入し窓の外に見えるプラズマを撮影した様子

53 295 3K 427K 563

View Details

Veritasium @veritasium

21 hours ago

Hackers Are Storing Data Now To Decrypt Later... #veritasium #science #physics #quantum

3 12 86 6K 9

View Details

Gary Marcus @GaryMarcus

17 hours ago

🚨Breaking news that could be huge, and enormously bad for GenAI, if other countries make similar decisions. the-decoder.com/landmark-germa…

25 100 301 38K 121

View Details

80 LEVEL @80Level

2 days ago

The PC version of the professional 2D animation app @callipeg, called Callipeg Studio, is coming later this year. It'll be available as a one-time purchase for €35: 80.lv/articles/2d-an… Video by @altamontagna170

0 6 60 7K 29

View Details

microbians @microbians

a day ago

@stats_feed Not same processing. Way far better.

0 0 0 62 0

View Details

La vida en viñetas @lavidaenvinetas

2 days ago

Hayao Miyazaki

1 81 707 8K 77

View Details

microbians @microbians

a day ago

Burro-c-rats

European Commission Audiovisual Service @EC_AVService

2 days ago

Apple decided not to roll out SIRI AI in the EU. The non-interoperability of their design made the update not suitable for the EU market. Why? Because big tech companies cannot decide which EU tool should be used by EU citizens. Learn more👇 📺youtube.com/shorts/ATg_W0N…

712 63 417 740K 244

0 0 1 47 0

View Details

𝔭𝔬𝔢𝔩𝔩𝔩𝔩 @poellll

2 days ago

school kids these days

3 41 309 48K 79

View Details

Andreas Storm @avstorm

2 days ago

Remember when people said Apple would never do that? Say hello to the Liquid Glass slider.

Andreas Storm @avstorm

11 months ago

Would be nice if we could decide how the glass should look in iOS 26

229 441 12K 537K 500

15 11 496 72K 57

View Details

microbians @microbians

2 days ago

@verge Too late? 😆

0 0 0 82 0

View Details

𝔭𝔬𝔢𝔩𝔩𝔩𝔩 @poellll

3 days ago

Katsuhiro Otomo

0 31 259 5K 21

View Details

Character Design References @CDReferences

3 days ago

Sherlock Hound (1984) #animationtidbits #sakuga #animation #2danimation

0 77 666 18K 172

View Details

Anime Rave @AniRave

3 days ago

This was animated 23yrs ago btw

46 301 5K 1.8M 3K

View Details

༺ཧคlคฝคཊ༻ @rybinfx

4 days ago

Variations 🐙 #touchdesigner

64 285 4K 158K 793

View Details

WebGL / WebGPU @webgl_webgpu

4 days ago

🧮 Physics that never leaves the graphics card. Kevin Levron's (@soju22) Three.js WebGPU demo runs the whole simulation in a compute pass, written in TSL, then draws thousands of bodies as one instanced mesh. 🔗 webgpu.com/showcase/three… #WebGPU #ThreeJS #TSL #ComputeShaders #CreativeCoding