Gustavo

@pontusab broooo

This is a big one, introducing `ai-sdk-tools/artifacts` for AI SDK. Build advanced AI interfaces with structured data and clean patterns. - useArtifact → receive streaming data and event from tools - artifact → register typed data schemas - createTypedContext → type-safe context across tools - Real-time updates with progress tracking

@cwd_harshit @orcdev @aisdk @vercel is that what makes all their templates laggy? Facing the same issue when using them as a base - even with no edits...

@orcdev @aisdk @vercel Bro they should not use react-syntax-highlighter, its slow and laggy and not optimized for re renders, they should use prism.js(directly)

Been playing last night with @aisdk 5 + AI Elements + AI Gateway by @vercel. Building AI apps has never felt this smooth. I’ll drop what I built soon ⚔️

@mattpocockuk Production ready agent patterns + performance tips too within NextJS. All demo chat applications seem to work fine but front end begins to lag heavily as you start adding tool calls etc

I think I know enough AI SDK to start cranking out tips What do you want to see? - API's you never knew existed? - Production-ready patterns? - Migration from v4 tips?

@MistralAI based

magnet:?xt=urn:btih:208b101a0f51514ecf285885a8b0f6fb1a1e4d7d&dn=mistral-7B-v0.1&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=https%3A%2F%https://t.co/HAadNvH1t0%3A443%2Fannounce RELEASE ab979f50d7d406ab8d0b07d09806c72c

@dannypostma lmao at the GA4 hate I felt the same initially but quickly realized it was objectively better (as a previous UA power user for years) or if you’re that desperate just use a LS template like this lol databloo.com/ga4-classic-in…

Huge startup opportunity to just literally rebuild GA3. Don't get creative, just 1:1 clone it.

@_akhaliq bloody incredible. Not lifelike just yet but not far off

Learning Physically Simulated Tennis Skills from Broadcast Videos demonstrate that system produces controllers for physically-simulated tennis players that can hit the incoming ball to target positions accurately using a diverse array of strokes (serves, forehands, and backhands), spins (topspins and slices), and playing styles (one/two-handed backhands, left/right-handed play). Overall, system can synthesize two physically simulated characters playing extended tennis rallies with simulated racket and ball dynamics paper: research.nvidia.com/labs/toronto-a… project page: research.nvidia.com/labs/toronto-a…

@_akhaliq “Winnie the Pooh cooking methamphetamine” 🔥

Mubert-Text-to-Music 🎵🎵🎵 Colab notebooks demonstrating prompt-based music generation via Mubert API GitHub: github.com/MubertAI/Muber…

@pess_r @EMostaque @runwayml This is beautiful

#stablediffusion text-to-image checkpoints are now available for research purposes upon request at github.com/CompVis/stable… Working on a more permissive release & inpainting checkpoints. Soon™ coming to @runwayml for text-to-video-editing

@EMostaque Twitter notifications on in case it reopens 🫡

And that's that, invites closed. Will see how it goes and reopen in a few days. Check #StableDiffusion tag for new creations.

🫡🫡