You're offline - Playing from downloaded podcasts
Back to All Episodes
Podcast Episode

NewsPodLM Daily Podcast 12 Jan 2026

January 12, 2026

Audio archived. Episodes older than 60 days are removed to save server storage. Story details remain below.

The B-side section covers rapid-fire updates on open-source reasoning models, compact high-performance AI systems, edge-optimised function calling models, production-ready audio-video generation, agentic AI for financial operations, and AI-native wearable devices. The episode emphasises the convergence of hardware innovation, open-source momentum, and commercial deployment across the AI landscape, highlighting how twenty twenty-six is shaping up to be a pivotal year for moving AI from experimental research to practical, cost-effective deployment at scale.

Topics Covered:
- NVIDIA Vera Rubin platform: CES twenty twenty-six unveiling of six-chip AI supercomputer with ten times lower inference costs
- Tiiny AI Pocket Lab: Three hundred gramme device running one hundred and twenty billion parameter models on-device with eighty gigabytes memory
- Moonshot AI funding: Five hundred million dollar Series C at four point three billion dollar valuation to build K3 frontier model
- Mistral 3 family: Six hundred and seventy-five billion parameter sparse mixture-of-experts and edge-optimised Ministral models under Apache licence
- Molecular computing breakthrough: Shape-shifting ruthenium complexes enabling memory, logic, and learning in single molecular structures
- World Labs Marble platform: Fei-Fei Li's spatial intelligence system generating exportable 3D environments from text and image prompts

Published January 12, 2026 at 6:46am

More Recent Episodes