“Near the zero point” — unstructured stream rambling about my excitement for N64 (M64) programming, demoscene & retro dev, LLMs / vibecoding, modern programming culture, and where useful / interesting work lies
Releasing the model weights and technical report of Kimi K3.
Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window.
New model architecture: 2.5x the intelligence per unit of compute, not just more params.
Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale.
Model weights: https://t.co/7m7eEg6Y0B
Tech report: https://t.co/yeu6cjpMCT
Tech blog: https://t.co/YTfiMSNM1f
Miyamoto was happy the Switch lasted as long as it did, commenting that specs aren't all customers want.
He also explained why you probably won't see Nintendo games on other platforms:
“First of all, I was happy that the Switch lasted for 8 to 9 years. I think that customers are no longer just looking for specs. I think the ideal is that ‘if you buy the software, you can play it on the device you have at home.’
Smartphones are ‘everywhere,’ aren’t they? I think that’s one ideal. But Nintendo has always been ‘providing gameplay that only we can create,’ including the interface, and compared to smartphones, it has the advantage of being able to respond quickly and having a high degree of freedom because it only requires one architecture and support for one piece of hardware.
If you use a general-purpose device, you have to support all versions, so the amount of work increases. We still want to make a living with minimal effort, to live without working… that’s half a joke, of course (laughs). But we believe that reducing the amount of work as much as possible is important for making development more efficient, so considering that, I think we’ve created a pretty good environment.”
(Machine Translation)
https://t.co/QIfg1Znskc
Just reverse engineered my Oura Ring 5 so I can control my computer like a wizzard. @ouraring please send my love to whoever buried a feature to stream live accelerometer data
@GMillar1874@acmilanchampo21 Y'all are going to love Atlanta Stadium (Mercedes-Benz Stadium). Right in the heart of downtown. You can ride the metro straight to it.
Today, we’re open-sourcing the draft specification for DESIGN.md, so it can be used across any tool or platform. We’re also adding new capabilities.
DESIGN.md lets you easily export and import your design rules from project to project. Instead of guessing intent, agents know exactly what a color is for and can even validate their choices against WCAG accessibility rules.
Watch David East break down this shared visual language in action👇. New capabilities and links in 🧵
This French Yu-Gi-Oh fan built a 3D system that makes cards summon monsters in real life. He developed it in just 7 months and built chips capable of reading more than 3,600 cards.
i trained Llama 2 on a $5 computer with less processing power than a thermostat.
it’s called 🥧 PiTorch: an ML library for Raspberry Pi Zeros.
i cover how to make it from scratch, from writing assembly GPU kernels to sending bytes over wires. (no hardware background needed!)
📏Lerp vs Slerp⭕
• Lerp interpolates in a straight line
• useful for interpolating positions
• Slerp rotates around (0,0)
• useful for interpolating directions
#unitytips
✅Functional
✅Modular
❌Pretty
✅Does the job!
Last week I cooked a small mission system, it can keep track of one objective or a whole bunch of them!
Nice to be back on BunnyBounce/SkaHop/BunnyGame/BootlegMario!
#gamedev#indiedev
My dear front-end developers (and anyone who’s interested in the future of interfaces):
I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at least in concept):
Fast, accurate and comprehensive userland text measurement algorithm in pure TypeScript, usable for laying out entire web pages without CSS, bypassing DOM measurements and reflow
Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc
Isaiah Rashad has a whole project called “Music 4r Da Vibers” not on streaming, and if you haven’t listened to it yet, go do yourself a favor and tap in