dadabots

we made the first fully ai synthesized album on bandcamp— it was black metal, anti-human, and unholy— now it’s officially banned art

🚀MIT Flow Matching and Diffusion Lecture 2026 Released (diffusion.csail.mit.edu)! We just released our new MIT 2026 course on flow matching and diffusion models! We teach the full stack of modern AI image, video, protein generators - theory and practice. We include: 📺 Videos: Step-by-step derivations. 📝 Notes: Mathematically self-contained lecture notes 💻 Coding: Hands-on exercises for every component We fully improved last years’ iteration and added new topics: latent spaces, diffusion transformers, building language models with discrete diffusion models. Everything is available here: diffusion.csail.mit.edu A huge thanks to Tommi Jaakkola for his support in making this class possible and Ashay Athalye (MIT SOUL) for the incredible production! Was fun to do this with @RShprints! #MachineLearning #GenerativeAI #MIT #DiffusionModels #AI

witness the weirdification of culture forking rapidly in parallel

Vibe coding is cool but have you tried vibe patching? Pure Vibes = Pure Data + MCP. Describe your sound, watch the patch appear 🌊 🔊👇

@varien i relate

how it feels tearing my meniscus playing meshuggah on drums

I hope you enjoy the flow of this playlist. Eclectic bass, trippy downtempo, IDM and more. Featuring my latest releases alongside favorite tracks from other artists. Updated every few weeks. Follow to stay in rotation. open.spotify.com/playlist/60ZxN…

@BasedBiohacker I took this for several months, on/off, low/med/high dose. Description feels exaggerated. Effect is more subtle, similar to piracetam. But YMMV, every brain is different. Caffeine has stronger cognitive benefits IMO.

FUCK i wish i could take the USSR cosmonaut stimulant phenylpiracetam more than once a week: - enhances acetylcholine utilization for learning consolidation and memory formation - working memory expands, holding multiple complex concepts simultaneously without losing bandwidth - pattern recognition becomes almost precognitive, seeing connections 3-4 steps ahead - 30-60x more potent than the OG nootropic piracetam - actually crosses the blood-brain barrier due to the phenyl group attachment - upregulates dopamine and norepinephrine receptor density in the striatum and prefrontal cortex - increases the number of receptors available to bind neurotransmitters, not just flooding them - verbal fluency improves, articulating complex technical concepts becomes effortless - processing speed increases dramatically - task-switching without context loss between complex problems feels effortless - abstract reasoning improves through sustained dopaminergic tone in prefrontal regions - actual stimulant properties, physical stamina and energy increase alongside cognition - anxiolytic effect at moderate doses, reduces stress response without sedation - enhanced mood and motivation through dopaminergic upregulation - neuroprotective, reduces cognitive impact of hypoxia and physical stress - hits in 30-45 minutes, lasts 4-6 hours - synergizes with noopept through completely non-overlapping mechanisms (my favorite combo) - anticonvulsant properties documented in russian literature - banned by WADA for physical performance enhancement (fuck yeah) - cold resistance documented in studies - developed for cosmonauts operating under extreme conditions - fat-soluble, take with dietary fat for maximum bioavailability unfortunately tolerance builds after a single use... dose max 1-2x/ week to keep the magic. original russian brand: cosmicnootropic.com/products/nanot… eu (ships from france): euro-nootropics.com/product/phenyl… us: kimerachems.co/product/phenyl… not medical advice.

@drawthingsapp MSL* not MLX

@drawthingsapp engineering.drawthings.ai/p/metal-flasha… ?

💥 When AI starts writing high-performance Metal shaders, everything changes! 🚀 LTX-2.3 VAE decoding is boosting. 🧪 Go try it yourself in the latest version of Draw Things.

is it just me or is the TL now a long list of vibecoded project launches. nice

YOOO royalcities just made a super version of stable audio open!! a finetune of the basemodel with a more deliberate dataset & prompting style

After months of work, today I’m releasing Foundation-1. A SOTA text-to-sample model built specifically for music production workflows. It may also be the most advanced AI sample generator currently available - open or closed. • ~7 GB VRAM • Entirely local • 100% free 😁

@deepfates too much corporate culture-setting has ruined everything

Who is actually setting AI culture right now? who defines the frame for everybody who's about to deal with this for the first time? do I have to start a podcast

Beyond language models and vision-language models, we observe similar effects in diffusion models. Here, one can think of “color thickets,” where certain regions of parameter space preferentially generate images with specific color palettes or visual styles.

@ViralManager @TravLauren dope

@dadabots @TravLauren hahaha were focused on the major commercial models rn i think the open source ai music crowd are just experimental artists and my beef is not with them lol

I'm looking for a super talented ML researcher who loves music. Dm me your experience.

@TravLauren @ViralManager detect our death metal vice.com/en/article/thi… youtube.com/live/JF2p0Hlg_…

@ViralManager The goats @dadabots can prob point you in the right direction

Open source must win.

if you are building at the edge of art and technology, we are hosting an artist residency in portugal in june alumni include @repligate @multimodalart @dadabots @exgenesis @blahah404 @AITechnoPagan @sares_at @neno_is_ooo dm if curious

Slop Studio Pro!!

@tmztmobile hahahaha yess thats the goal

@dadabots Thunderbolt in/out stompbox?

Fuck yeah. New project is called SPEED METAL. We just achieved 9ms latency on M4 metal gpu w/ our new streamable neural decoder using MLX. Is this beatbox-voice-changer levels of speed? :O