Okay, Fable 5.1 is crazy good at creating three.js sites. It’s faster, understands complex design instructions better, and recreates references with surgical precision. The English is also much easier to read. With this much power, it’s hard to settle for static sites, especially when so many AI-made sites look generic. A few things I noticed though: - It still creates generic AI illustrations if you don’t specify the images. - Difficult 3D subjects like people and dogs still become 3D slop unless I use Higgsfield 3D models. - You still have to push it to improve the textures, lighting, modeling, and small details. - Taste still matters. I had to fix overlapping elements, negative space, and scroll behavior. - I used and for the particle, VHS, and CRT effects. - Because it works faster, I go through tokens like there’s no tomorrow. Inspired by I can open-source this if anyone is interested.

Okay, all my initial benchmark tests with Fable 5.1 in @VulcanBench worked, so I can now run the full sweep. As someone that benchmarks models across all effort levels, and that just raised the timeout from 2 hours to 10 hours, yeah, this is going to take a while. But it's the only way I think we can really do this right and hopefully answer the question: as an engineer, or engineering leader, trying to make sense of when to use the latest frontier model for coding, and at what effort level, what should I do. I hope to help you answer that questions...in 90 - 120 hours.

Okay… some context is definitely needed here 😂 Today we gave the MAME community creative control over one meme. We started with MAME on a completely blank background. Then, round by round, the community decided everything: 📍 Where? — The Oval Office. 🕺 What is he doing? — Dancing with Trump. 💵 What is he holding? — Cash. 🦩 Final detail? — A giant inflatable flamingo bouncing around and destroying the office. You decided. We created. And somehow… this is what we ended up with. 😂 Our first Community Meme, made entirely from your ideas. 🖤🐾 Did we have fun? Definitely. Is it weird? Absolutely. Do we regret it? NEVER😂 Not if a little weirdness can bring this much fun and positive energy into our community ⚡️

Okay, I can now proudly say I have been a Muse Code Power User from day one, excited! If you haven't seen the benchmark I did with @VulcanBench last week in Pi vs. a bare bones harness, you can check it out here: Kinda wild, but it was my own benchmark that convinced me to get this plan. And yes, I will probably end up using every token in the plan just benchmarking the Muse Code harness, but maybe there will be some tokens left for me to play with too! 🤞