David

Mac Mini M5 vs RTX Pro 6000, which one is faster for llm inference?

@kk_design_dev good quality... can you show how to do that?

実際のところゲームどうなんだろうとOpus4.8への指示でやってみた ClaudeCode経由で実装、音声、画像、発話、モデリング全部やってくれる FableでなくOpusだけで充分ありがてぇ！ 楽曲・効果音生成でOSSはずーっとイマイチな印象だったけど楽曲生成でのAceStep1.5、効果音生成でのStableAudio3がめっちゃ良くなってた ●ゲームエンジン : Rust Bevy Engine ●ComfyUI 楽曲 :　Ace Step1.5 効果音 :　StableAudio3 画像、テクスチャ等 : Qwen Image 実況音声 :　Vertex Gemini TTS → Irodori-TTS ●3Dモデル ・BlenderMCP→Sketchfab ・ComfyUI Qwen & Trellis2 → BlenderMCP ・ComfyUI Qwen & Hunyun3D → BlenderMCP

what would you do if you were given a final chance to use Fable 5 free for 24 hours? what would you use it for?

The US gov banned Fable, but I doubt they’ll stop the incoming wave of same level or more powerful models.

@nuvolore Cool

@tvcdave No, you can run it locally, it has a CLI and Skills

Yesterday, before they took down Fable 5, I was able to build a 3D editor tool that turns text into low poly 3D assets for games. It does work also with Opus 4.8 and GPT 5.5 but obviously Fable 5 was a bit more mindblowing. This doesn't use technology like Meshy or Tripo. Each model is built with a STEP file format, which LLMs understand well and it also allows for super easy editing, like: "Make the countertop bigger". The second part, was solving a fast way to texture those assets using AI image generation, combined with texture projection. In the 4 images here there: 1) Initial concept art. 2) Concept transformed into a CAD sheet. 3) A rendered image at a specific camera angle. 4) Text-to-3D with STEP and Texture projection. The interesting unlock that game developers should fully embrace is that AI makes it much easier to create custom tools for exactly what your game needs. In the past, a developer had to jump between many different tools, each with its own workflow, limitations, and learning curve. Now you could build a focused tool in a day, use it immediately, and keep adding features only when the project actually needs them. Not every tool needs to be a product. Not every pipeline needs to be universal. Sometimes the best tool is the one built quickly for a very specific job.

@adamkhootrader Can I sit in?

Spending Sunday morning coaching daughter on analyzing charts , advanced technical analysis She’s up 54.77% last 12 months but still not happy, asked for private coaching

@nuvolore Nice work

@itsshara_ai @ReelQuestAI Can it create RPG, SLG and platform games as well?

I built a full action sports fantasy game using only AI. @ReelQuestAI turns any idea into a fully playable interactive game, no coding, no dev team needed. Just imagine it. AI builds it. Create now: reelquest.ai/go/tw_us_shara

@maverick23NFT We need to have our own AI and the ability to develop it and the infrastructure. So we don’t rely on others. What if others decide to not offer their service? Or raise a high price?

Singapore doesn’t lack capital, talent or infrastructure what it lacks is fortitude and belief in home teams hence why SG tend to end up as investoors and managers but never owners & inventors

Do we have any plan to develop our own frontier model, so that we have more control of it? @LawrenceWongST

We have : 1. Prompt engineering 2. Context engineering 3. Harness engineering And now.. loop engineering

My new project

Is openclaw really useful? I use Claude code in VS code. I don’t understand why people use OpenClaw for coding or building app, because it seems to break down often and … it is not an ide.

@karpathy @bcherny Wow

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

@YashHustle_22 I am using cc, works pretty well, except the rate limit.

Anyone here still fully loyal to Claude Code? Most devs around me seem to be experimenting with Codex now.

Why Anthropic doesn’t have an auto mode option which auto select model depending on complexity of the tasks? This should be able to reduce the tokens consumption, which many users are facing. @bcherny

📈👀

My biggest issue with Opus 4.7 on Claude web: Only “Adaptive” or non-thinking. No way to force thinking mode. And it doesn’t even know Opus 4.6 exists, and I cannot force it to think and do web search mid conversation!

@forgebitz Gemini 3.1 pro with antigravity is good, until the Google lowers the weekly quota limit and reset.

i don't know a single developer using gemini models