an abstracted 3d representation of the world is def cool too — depending on the application
MULTIMODAL AI
-
Preserving Figures in AI Chat: A Solution for Scientific Papers
By
–
When you ask ChatGPT, Claude, Gemini about a paper, they strip out the figures and give you prose. Scientists choose to include plots and diagrams for a reason — they are dense and precise. I've cooked a way to preserve them in our chat agent, with architecture diagrams, tables, plots, linked back to the source PDF. How was this not already a thing? tiptreesystems.com/blog/comp…
→ View original post on X — @hugo_larochelle, 2026-03-17 17:50 UTC
-
Devil’s Advocate Podcast: AI Characters Discuss News Live
By
–
This can go in a lot of directions! https://t.co/rlR9coH3XE
— Gene Kogan (@genekogan) 17 mars 2026This can go in a lot of directions! nitter.net/c_valenzuelab/status/2… Cristóbal Valenzuela (@c_valenzuelab) Introducing the Devil's Advocate: a live podcast where God and the Devil discuss current events and news in real time. I connected two @runwayml characters to each other and had them speak and discuss the news live. The Devil is going through a rebrand and is deep into self-help, and God is a bit tired and hasn't slept in a while. Thinking of adding a mode where they can take calls from the audience. We could also technically run this 24/7. — https://nitter.net/c_valenzuelab/status/2031474004956225786#m
→ View original post on X — @genekogan, 2026-03-17 17:29 UTC
-
Avatar Duet: AI Conversation Loop with Runway Integration
By
–
You can try this with a Runway key here: avatar-duet.vercel.app/ It's simple, just routes each other's audio to the other instead of the mic. Code link in the page. I'm wondering if they'll eventually find that "spiritual bliss attractor" state if they talk long enough.
→ View original post on X — @genekogan, 2026-03-17 17:27 UTC
-
Runway Live Avatars Conversing in Browser Demo
By
–
Having some fun making @runwayml Live Avatars talk to each other in the browser. pic.twitter.com/DrSQsEkFzA
— Gene Kogan (@genekogan) 17 mars 2026Having some fun making @runwayml Live Avatars talk to each other in the browser.
→ View original post on X — @genekogan, 2026-03-17 17:21 UTC
-
Google Street View Aerial Data Powers Grounded AI Genie
By
–
it's all over when google realizes the treasure trove that is street view + aerial data and launches a version of genie grounded in the real world… https://t.co/CWQ0vPjNuJ
— Bilawal Sidhu (@bilawalsidhu) 17 mars 2026it's all over when google realizes the treasure trove that is street view + aerial data and launches a version of genie grounded in the real world…
-
OpenAI Launches GPT-5.4 mini, Twice as Fast
By
–
GPT-5.4 mini is available today in ChatGPT, Codex, and the API. Optimized for coding, computer use, multimodal understanding, and subagents. And it's 2x faster than GPT-5 mini. https://openai.com/index/introducing-gpt-5-4-mini-and-nano/ [Translated from EN to English]
-
French Synthetic Data at Scale: Regional Variation Challenges
By
–
Super cool to see French-grounded synthetic data at this scale! As a French-Canadian, curious how the demographic distributions handle regional French variations 🙂 It's usually pretty bad especially voice generation haha.
-
RULE Framework Tackles Noisy Multimodal Entity Alignment
By
–
Is inconsistent data making multi-modal entity alignment a nightmare? Researchers at Sichuan University, China, have a breakthrough! They introduce RULE, a framework designed to conquer "Dual-level Noisy Correspondence" (DNC). RULE intelligently assesses the reliability of
-
Apple and Google sign $1 billion deal to power Siri with Gemini
By
–
Apple and Google quietly confirmed a deal worth ~$1 billion/year. Gemini will power the new Siri — with on-screen awareness, cross-app intelligence, and real context. The world's most-used AI assistant is about to get serious. applemagazine.com/apple-targ… #AI
→ View original post on X — @svenphilipsen, 2026-03-17 10:00 UTC