Here’s my very weird observation that shouldn’t make sense but is correct: this tweet is right about AI books and movies. But AI generated music has very real traction and good ones are listened to on repeat a lot, the same user behavior as normal hit music. Matteo Pellegrini (@matteopelleg) Nobody wants to read AI-generated books, watch an AI-generated movie or listen to an AI-generated song. — https://nitter.net/matteopelleg/status/2037546434183180590#m
Here's why I shill Droid 24/7 ———- Today Droid single-handedly: 1. Published a REAP of GLM-5 in FP8, there's a reason no one else has done it DSA is still very new: huggingface.co/0xSero/GLM-5-… 2. Found and Fixed an upstream issue with VLLM + DSA + Hopper where GLM-5's kv-cache would need to recompute and spend 20x the time needed, fixed. 3. Created multiple working quantisations on it's own, it tried exl3 and autoround but both failed so resorted to GGUF (autoround 3 bits doesn't work on ampere) huggingface.co/0xSero/GLM-5-… 4. Implemented github.com/0xSero/turboquant within 24 hours of the research paper coming out, tested it across 5090s, 3090s, H100s, and B200s 5. Has been distilling larger models into LoRA to help me test arxiv.org/abs/2505.21835 and it got an 80% prune to be semi-coherent again. 6. Helped my find research papers, clean up slop with the human-writing skill. 7. Got BYOK working with Anthropic, ZAI, Kimi, MiniMax, OpenAI working in Cursor github.com/0xSero/factory-cu… 8. Helped me Implement blog.comfy.org/p/dynamic-vra… 's dynamic loading, only works on a tiny model, but still. ——- I only have to check in on it every 30-45 minutes (I am talking all 8 of my sessions) the thing will run for 16 hours with like 0 prep All this while I am mostly focused on my actual job and tweeting 24/7 Keep in mind each one of these experiments is running on a different server, with different constraints, like I don't understand how I can get such good results here. ——— I love novelty. Which is why I jump around and talking about all these different tools. I have used all of these harnesses and messed around with every feature. I keep coming back to this, and I keep shilling it because I sincerely wish others get to experience this.
one of my favourite plugins in codex is Build Web Apps, it combines @shadcn & react best practices with web design guidelines! all of it with the ability to deploy on Vercel and connect to stripe & superbase you can literally build a startup with just this one plugin!
5/
What LTX seems to be doing is positioning itself less as a single model and more as a workspace for this kind of process. If you’re working on concepts, pitches, or early-stage visuals, it’s worth trying:
3/ That reliability is the key part. In most AI video tools, you spend a lot of time regenerating and fixing things that are "almost right." If that loop gets shorter, the tool becomes much more practical.
2/ What stands out in this release is not one flashy feature, but the combination: – native 4K portrait video
– keyframe control
– much better prompt understanding
– cleaner audio
– fewer unusable generations
1/ AI video is getting crowded. What matters now isn’t just who can generate a cool clip, it’s which tools actually fit real production workflows. LTX-2.3 inside LTX Studio feels like a meaningful step in that direction. Open weights and local runs are impressive, but LTX
Microsoft -25.9% in Q1 2026. Besides Meta, it's the only company that hasn't managed to properly integrate AI into its processes. And apparently, not even Azure can salvage expectations.
