The Daily AI Drop - Evening Edition, November 1, 2026

A looped multimodal model you can run locally, the trillion-dollar fight over who controls enterprise AI agents, and why you should write down your boldest ideas.

Listen: Listen to this episode

SenseNova Loop 9B: Hands-On Test of a Looped Multimodal Model — fahdmirza

• Watch source video ↗
• Fahd Mirza locally installs and tests SenseNova-Loop-9B-Preview, a nine-billion-parameter multimodal model that takes in text and images.
• Its looped architecture is built for long-context and agentic workflows, covering document understanding, coding, and external tool use.
• That a model this capable is runnable on local hardware matters: it keeps your data under your control and turns agentic workflows from a hosted service into something you can inspect and run yourself.

Salesforce Just Helped Microsoft. You're Picking The Winner Of A Trillion-Dollar Fight. — NateBJones

• Watch source video ↗
• Nate B. Jones examines why Salesforce helping Microsoft's agents, and Oracle bringing its applications into ChatGPT, are moves in a fight over who coordinates work across the software a business already owns.
• The prize is the agentic control plane: the layer that defines which agents can touch which data and what they are allowed to do.
• For anyone choosing AI tools today, interoperability is the thing to watch, because once a control plane is entrenched, switching ecosystems gets very hard.

The "Ridiculous" Idea You're Afraid to Write Down — aishwaryasrinivasan

• Watch source video ↗
• Aishwarya Srinivasan makes the case for writing down the ambitious ideas you are tempted to dismiss as ridiculous before they ever leave your head.
• In a field moving as fast as AI, today's ridiculous idea is often tomorrow's product, and the builders who win are frequently just the ones who take an odd idea seriously a little longer.
• The practical takeaway is simple: capture the idea first and judge it later, because an idea thrown away in your head can never be developed.