The open-source AI revolution is shattering the boundaries of enterprise customization and mobile hardware limits. Mira Murati’s highly anticipated startup, Thinking Machines Lab, has officially entered the arena with "Inkling", a staggering 975-billion parameter Mixture-of-Experts (MoE) model built to compete directly against tech giants through bespoke business fine-tuning.

We’ll talk about:

Mira Murati's startup launches a 975B parameter MoE architecture that only awakens 41B active parameters per task, featuring an adjustable "thinking effort" dial and native uncertainty flagging.
How Thinking Machines plans to monetize by using Inkling as a baseline for enterprises to build hyper-customized, private-data workflows using their training hub.
Bonsai 27B utilizing aggressive 1-bit quantization to shrink a 27B model down to an unprecedented 3.9GB, running local coding and vision workflows entirely on-device without cloud costs.
OpenAI launching "Codex Micro," a dedicated physical keyboard featuring built-in workflow dials, while Apple reportedly taps Alibaba's Qwen to finally power Apple Intelligence in China.

Keywords: Thinking Machines Lab, Inkling, PrismML Bonsai 27B, phone local LLM, Codex Micro, Tinker.

Links:

Newsletter: Sign up for our FREE daily newsletter.
Our Community: Get 3-level AI tutorials across industries.
Join AI Fire Academy: 700+ advanced AI workflows ($14,500+ Value)

Our Socials:

Facebook Group: Join 295K+ AI builders
X (Twitter): Follow us for daily AI drops
YouTube: Watch AI walkthroughs & tutorials

🎙️ EP 313: Thinking Machines Drops "Inkling" & PrismML Fits a 27B Model on Your Phone

AI Fire Daily

🎙️ EP 313: Thinking Machines Drops "Inkling" & PrismML Fits a 27B Model on Your Phone