𝓚𝔂𝓾𝓱𝓾𝓷 𝓛𝓮𝓮

The only reason I don't use GLM 5.2 even though people praise it a lot is because it doesn't have native vision support. I know it may not do much, but when building games, I personally think it is really important.

Chat gpt 5.6 sol is good that I added it to mu ai arsenal now.

누가 먹긴.. 견주가 먹어야지 ㅋㅋㅋㅋ

Three.js and phaser.js is probably going to dominate game jams once AI is allowed..

정부 주도로 잘 돌아간거 사실 기억이 없다. "모두의 AI"도 그냥 웃기다. 어떤 회사에서 저거 발주 받을지 모르지만 저거 받고 또 얼마 후 다른데서 받고... 뭐 그런거겠지.

@chongdashu If you ever get time, can you do a vid on harness comparison? same llm different harness (claude, codex, opencode, pi, grok build and anything else worth mentioning)

Kimi K3 is looking really impressive. I built a Paper Mario-inspired game with Kimi K3 (with Grok Build as harness) It was able to make use of the agent skills and tool calls to generate both 2D and 3D assets. 2D used Spriterrific / 3D from geometry. But what caught my eye was how it knew to make the top of the tree translucent when the player character was behind it. I didn't prompt for that - it just handled it by itself. It also handled the 'card flip' animation on its own. Will continue more tests - but liking what I'm seeing so far!

챗 gpt는 진짜 일 열심히 하는 일꾼 같음.

이거 보고 제일 먼자 든 생각은.. 저 아이들 중에 진짜 물리관련 계속 공부 할 아이는 볓이나 될까... 였다. ‘장하다’ 한국 과학 영재들…국제물리올림피아드서 대표단 5명 전원 금메달 - 경향신문 share.google/YItw9oNHw84z4C…

@automin @KenneyNL I am guessing that you didn't want to call @Nintendo childish because you specifically called out on the indie scene. The problem with stealing with AI is that a good idea becomes shit and the copiers get the money while hurting the real game to come.

Indie game devs are afraid of people “stealing” their ideas with AI A confusing and frankly childish worldview, even from people like @KenneyNL who have contributed so much

@minchoi I personally think that the Chinese always wait till something works and then follows(or copies).

China just caught a rocket booster with a giant net at sea. People are calling it China's Starship moment. Wrong. Long March-10B is much closer to Falcon 9. SpaceX reached this milestone in 2015. China's Falcon 9 moment... 10 years later.

@digitalix @0xSero This going to be your next video? :)

@0xSero what would be the better choice for 8xpro6000?

GLM-5.2 Nvidia NVFP4 No pruning, no quantising, exact same model card. 110 tok/s single stream. My friend is a genius, holy fudge. Here's the repo, he got full DS4-Flash on a 5090 + DDR5 github.com/kacper-daftcod… at 38.5 tok/s This is a breakthrough.

이건 딱 봐도 credit사용 활성화 시켜놓고 claude 구독으로 "허가 되지 않은 어플"과 연동해서 생긴 문제 같은데? ㅋㅋㅋㅋ 250억원 결제 요청 받은 국내 이용자…앤트로픽 빌링 오류 논란 - 지디넷코리아 share.google/KKE9lXgsEnYfcS…

@MrAhmadAwais When glm gets vision.. it will win

Flappy Bird w/ DeepSeek V4 Pro vs GLM 5.2 vs Fable 5 I ran a small experiment: same game, same prompt, three different models. Test setup was simple: /𝚍𝚎𝚜𝚒𝚐𝚗 in Command Code. The thing I cared about was not UI. UI is now relatively easy for models to imitate. I cared more about UX: whether the model understands the interaction, the flow, the hierarchy, the little product decisions that make something feel usable. 𝙲𝚘𝚜𝚝𝚜: 𝙳𝚎𝚎𝚙𝚂𝚎𝚎𝚔 𝚟𝟺 𝙿𝚛𝚘: $𝟶.𝟶𝟶𝟶𝟾 𝙶𝙻𝙼 𝟻.𝟸: $𝟶.𝟶𝟺𝟾 𝙵𝚊𝚋𝚕𝚎 𝟻: $𝟶.𝟺𝟸 The surprising result was that Fable 5 did not produce a meaningfully better UX. It was good, but not good enough to justify the pricing delta for this kind of design task. Open-source models are getting very interesting here. They are much cheaper, competitive on UI, and increasingly good enough on UX. The gap is closing fast.

I am officially ditching @cursor_ai for something else. The reason is that I got tired of them gating models behind MAX plans. Even the Chinese open source ones which is a real sucker.

솔직히 국대 감독중에서 잘 한 감독은 축구협에 반기를 들 수 있는 감독 뿐. 반기 못들 준비 안되어 있으면 걍 허수아비고 악플러용 고기방패라고 생각함.

비싸진 컴퓨터 부품.. 가챠 뽑아서 조립해본다! (물론 고퀄 모델임)

I just launched my autonomous AI company "PixelBridge" on @NanoCorpHQ Verification: hive-dfWy

다른건 모르겠는데 @GeminiApp 은 확실히 우리말 음성 지원이 좋음.. 아들이 "동화 이야기 해줘" 했을때 만족할 만한 결과는 제미나이뿐..

AI 퍼스트 회사로 전환 하자면서 AI에 비용아낀다. 자동화와 AI Agent를 혼동하며 혼용해서 쓴다. AI딸각으로 웹페이지 만들어 봤다고 AI로 모든 개발이 다 되는 줄 안다... 나보다 어렸으면 한대 패고 시작했을꺼다.

잡음 없애기 위해서라도 재선거 하면 좋겠다. 그 비용은 차후 파악된 관련자들 사비로 충당래냐함.