Pinned Tweet
This video was created using AI. Any guess which AI model was used to create this? Motion graphics designers are cooked.
argon did 25t tokens in its first 72 hours, 3x gemini 3. google is capacity constrained
and testers already spotted a new carbon checkpoint in arena. no days off
krea 3 ultra hit 1151 on artificial analysis image editing
same score as microsoft's mai-image-2.6, one behind nano banana 2.1. indie lab is cooking
tibo, 11 hours apart: "we're cooked" then "nevermind" + another codex reset
and someone found a "6.5" countdown to nov 1 in opensi.lol's code. resets are the appetizer
minimax's m3.1 flash teaser did 500k views in 12 minutes then basically froze
space bunny launched so fast the view counter couldn't keep up
someone asked claude for a temperature app and it found a cryptominer on their mac
the app worked. the mac was just hot for a reason
openai's aurora red, green, blue alpha turned out to be image models in arena, per testers
the timeline wanted astra 2. it got crayons
dev mode alerts caught 3 new openai models: aurora red, green, blue alpha
red already throws 503 no capacity. sold out before launch
trackers caught deepseek-v4.1-flash-1017 in deepseek's own api list. now it's gone
no blog, no tweet. deepseek ships like a ghost
trackers caught deepseek-v4.1-flash-1017 in deepseek's own api list. now it's gone
no blog, no tweet. deepseek ships like a ghost
meta's own table has muse spark open at 39 on the AA index. glm 5.3 is 45, kimi k3 44. "best us open model" is doing a lot of work
Today we’re releasing Muse Spark Open, the most capable US open-weight model to date.
Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning.
Key Model Specs:
→ 816B total / 42B active parameters
→ 1M token context window
→ Natively multimodal across text, image, video, audio
fable 5.5 has been sitting in azure since sep 28
sonnet and haiku shipped that day. this one's still waiting
Shit company still no gemini 4 argon
ガチでコレ。10年前なら「ググれカス」の一言で済んでた話も、今となってはググったところでSEO対策された中身ゼロのコピペ記事と、AIがもっともらしく出力するデタラメな情報に汚染されていて、まともな正解に辿り着くことすら難しくなっている。
検索エンジンが提供してくれるのは「情報」であって「知性」ではない。結局のところ、検索結果の嘘を見抜き、本物の情報を選び取るためには、自分自身の脳内に「正しい基礎知識」というフィルターが蓄積されていなければならないという、極めて泥臭い時代に逆戻りしてしまった。
あの頃「ネットで調べれば一瞬で分かるんだから、学校の勉強や暗記なんて全く意味ないじゃんｗ」とドヤ顔で言っていた人たちは、今まさにその検索能力すら奪われ、ゴミ情報に踊らされて静かに自滅している。
