Nevv🗿

@kilocode @XiaomiMiMo What's the deadline

🚨 Virtual Hackathon Alert: Kilo x @XiaomiMiMo Last month, we asked developers to build the worst website ever. More than 500 of you showed up. Now we're back with a new, free hackathon: Build the WORST captcha experience ever. Prove you're not a robot by screaming into your mic. Calling your mom. Solving a mystery. Whatever it takes, get creative. $1,250 in prizes is on the line, starting now!

So many startups think their engineers are "cracked" but have no idea what that really means. This team of 5 19yr olds built a 30 petabyte storage cluster in SF for ~$500k to get a 40x cheaper AWS S3 as a side quest to store 90M hours of video. Now, that's cracked.

Forcing kids to vegetate for 12 years when they could play the entire curriculum like an RPG tree at their own pace is by far the greatest instance of child abuse we've passively accepted as ordinary.

@Stefan_3D_AI Rigging for custom models that are not characters still sucks btw

AI is not “coming” for 3D. It’s already here. 📍Base mesh? Automated. 📍Retopo? Minutes. 📍Rigging? One click. 📍Mocap from phone video? Done. The real threat isn’t AI. It’s staying at the execution level while everything automated. Watch the full breakdown: youtu.be/19qOD7ryN2U #3DAI #3DArtist #WorkflowShift #AIFuture #3DModeling #GenerativeAI

@GZinMetaverse @fdotinc Possible to participate online? 👀

Everyone knows world models are the future. That’s why we’re bringing another build weekend(Mar 14-15) to @fdotinc , where physical AI builders ship world models across XR, gaming, and WebSpatial. Call it Episode One. The physical world will come next. If you want to be part of the series — start here👇 luma.com/worldsinaction…

@RadiantsDAO 👀

The last comment gets the $10,000 Hackathon Prize 👇

@SeloSlav @spacetime_db What did you use for terrain generation?

Built a procedural open world with @spacetime_db as the backend. Every terrain chunk, tree, and stone is server-authoritative. The client subscribes and the world just appears, consistent across every connected player, no REST endpoints, no polling, no manual sync logic. The database is the game server. Now to fill it with agents!

@taylor_sntx Where are you getting heightmap data from?

hologram map working with live heightmap data from the earth... here I'm exploring the Himalayas, Grand Canyon, and Andes mountains. it accepts any longitude/latitude and then you can pan around. works in the browser powered by three.js

you basically need to be unemployed rn to keep up

@pixelvaultst Link?

Free PNG pack, over 200 assets! ✨

@AlexFinn "Can run on almost any modern computer" "If you have 32gb of RAM" 😐

Do you even understand what this means? An open source model just released that is: • Just as smart as Sonnet 4.5 • Incredible at coding • Can run on almost any modern computer If you have 32gb of RAM (most Mac Minis do) you can have unlimited super intelligence on your desk. For free. Sonnet 4.5 was released 5 months ago In 5 months that level of intelligence went from frontier to free on your desk And not only that, can run on any laptop with 32gb of RAM If you have the memory, do the following immediately: 1. Download LM Studio 2. Go to your OpenClaw and ask which of these new Qwen models is best for your hardware 3. Have it walk you through downloading and loading it 4. Build apps with it knowing you are using your own personal, private super intelligence on your desk The people denying this is the future are so beyond lost.

Reinforcement Learning 10,000x Faster - an invited talk for the University of Warwick's AI summit. I share the story of Neural MMO and how a crazy out-there MMO AI project became the basis for PufferLib, the fastest RL library available today. Star puffer on github to support!

@Iuvnriki @magicblock The one I submitted is 😇

@magicblock how many of these projects are actually playable?

30 projects submitted to the Solana Blitz v0 Hackathon Games, DeFi, privacy, infra, and more - all built with MagicBlock Winners dropping soon 👀

Fully onchain gaming is the future

@heettike Public hackathon wen

we're doing an internal hackathon later this week & might just tokenise all 5 projects with noice age of abundance for software, agents & tokens

@gregpr07 Too bad it's not in Europe 😩

The three biggest names in AI just signed on to sponsor one hackathon 👀 Google Deepmind, Anthropic, and OpenAI are now on board for the world's largest web agents hackathon. $150K+ in prizes. Don't miss it. 🌁 Register below ⬇️

@taylor_sntx Did you make source code for it public?

i want visualizations to feel more organic, less sharp and perfect. like a well-worn hologram. this three.js visualization uses a few tricks - particles arranged in rings instead of a grid, variable density that decreases with height, and falloff opacity

out of necessity, the internet will soon be made up of geofenced, human-verified, invite-only micro-communities

@askOkara Honestly tho, opus still generates higher quality code in my experience

dude, minimax m2.5 is insane. i gave the same prompt to both opus 4.6 and minimax 2.5, and they built the same thing, but minimax is like 20x cheaper than opus 4.6. the first version is from opus 4.6 and the second is from minimax m2.5.

@LiorOnAI Wanted to start with BCI, any idea what are currently cheapest affordable hardware options?

You can now turn cheap EEG headsets into lab-grade brain scanners. And it's open-source. ZUNA is a 380M-parameter foundation model that reconstructs missing brain signals from partial EEG data. It works across any electrode setup, from consumer headsets to 256-channel research systems, without retraining. It lets you: - Reconstruct missing EEG channels from sparse data - Denoise corrupted signals - Predict new channels from just electrode coordinates - Handle arbitrary electrode layouts The model uses a diffusion autoencoder with a transformer backbone. It was trained on 2 million channel-hours across 208 datasets using masked diffusion training and 4D spatial embeddings. This lets the model understand the physical geometry of electrode placement. Each channel signal gets compressed into tokens, then the model encodes x, y, z positions plus time into separate attention components. EEG data has been stuck in a pre-foundation model era. Datasets are small, fragmented across institutions, collected under different protocols. The standard fix for missing channels is spherical spline interpolation, basically spatial smoothing. It works okay when a few channels drop out but falls apart when you lose more than 75% of your data. ZUNA beats this baseline by learning actual patterns in brain activity instead of just smoothing between points. The gap widens dramatically at high dropout rates, exactly where you need it most. Thought-to-text is positioning itself as the next major AI modality after language, vision, and audio. But you can't build that future on data that gets thrown away because a few electrodes failed. The model is fully open source under Apache 2.0, runs on consumer GPUs, and works on CPU for many tasks.