Home
Welcome
About
Profile & credentials
The Casebook
Real case studies with measured outcomes
Product Playground
Try a concept. Play with a product.
The Radar
Insights on AI in logistics
AIHot
Real-time AI news feed
The Manifest
Daily AI × supply chain signal
October 6, 2026 · 24 stories

The Manifest

Daily · AI × Supply Chain

CH Robinson's $5.8B purchase of RXO reshapes freight brokerage just as Suez traffic returns to squeeze transpac rates, while the AI market keeps shuffling, DeepSeek nearing a $12B round, Reflection AI shipping an open-weight rival, and Anthropic's big customers pulling back.

Get The Manifest in your inbox

One sharp digest of AI × supply chain. Free. No noise.

Today's Top

  1. 01CH Robinson to buy RXO for $5.8B, combining 3PL heavyweightsSupply Chain Dive
  2. 02CATL and Tencent back DeepSeek's ballooning funding round as the AI startup eyes a 2027 IPOThe Decoder
  3. 03Meta and Microsoft pull back from Claude as Anthropic transforms from partner into competitorThe Decoder
  4. 04Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute costTechCrunch AI
  5. 05Suez return a major threat to carrier efforts to halt transpac spot rate slideThe Loadstar
01

Models & Releases

5 stories

Cohere pitches North 2 as the enterprise AI control room that works with any model

This is Cohere betting enterprises want an orchestration layer rather than another chat model. Worth a look if you're juggling multiple vendor models and want one place to manage agent workflows instead of stitching it together yourself.

The Decoder↗ source

Reka Releases Rho-1: A 19B Omni-Reasoning Model That Understands, Generates Video and Outputs Robot Actions in One

One network for text, video and robot actions is a research preview, not a shipping product, so treat the robotics claims as early. Still a signal that the line between planning software and physical automation keeps blurring.

MarkTechPost↗ source

Meet Together Link: A Free CLI That Runs Open Models Like Kimi K3 and GLM 5.3 Inside Claude Code, Codex, and OpenCode

A free router that swaps cheaper open models into your existing coding tools is exactly the kind of cost lever procurement teams should be tracking as AI tooling spend grows line by line.

MarkTechPost↗ source

Aleph Alpha releases Kolibri, an open-weight model that makes the case for European AI sovereignty

Another open-weight entrant aimed squarely at buyers who need data residency and language coverage outside the US and China options. Good one to keep on the vendor shortlist for EU-based operations.

The Decoder↗ source

OpenAI will watermark ChatGPT text in the EU but makes it optional for API users worldwide

Detection rates drop fast once text gets edited, so don't treat this as a reliable compliance tool. Enterprises relying on AI-generated content for EU markets should track this closely before it becomes an audit requirement.

The Decoder↗ source
02

Supply Chain & Ops

5 stories
03

Deals & Market

4 stories
04

Research & Frontier

5 stories

MCP for agent-to-agent comms may be the riskiest protocol you've never heard of

If your vendors are wiring agents together over MCP, ask how they're containing prompt injection between agents. This is a structural flaw, not a patchable bug, so expect more incidents before it gets fixed properly.

Ars Technica AI↗ source

Beyond Domain-Specific World Models: JEPA-Anything Uses 1 Recipe for 7 Fields

A single world-model recipe generalizing across seven domains is the kind of research that eventually trickles into forecasting and simulation tools. Not operational yet, but worth bookmarking.

MarkTechPost↗ source

Building a Streaming Robotics Learning Pipeline Using NVIDIA Cosmos3-DROID

A practical walkthrough for teams experimenting with robot learning without standing up local storage infrastructure first. Useful if you're evaluating warehouse robotics pilots on a budget.

MarkTechPost↗ source

Yandex Introduces Sona: A Single Generative Recommender That Replaces Entire Recommendation Cascade

Collapsing a multi-stage recommendation pipeline into one transformer and still beating the old cascade is a pattern worth watching for anyone running demand sensing or assortment tools built on legacy ranking stacks.

MarkTechPost↗ source

Researchers are tracking a Chinese AI 'agent fleet'

An autonomous agent swarm probing a rival's map service is a preview of what automated competitive intelligence and scraping look like at scale. Expect more of this kind of agent-versus-agent activity to surface.

TechCrunch AI↗ source
05

Org & AI Architecture

5 stories

Most Americans want AI development to slow down or stop entirely, new poll finds

77 percent wanting a pause and 86 percent wanting independent safety standards is a trust gap that eventually becomes a regulatory one. Leaders rolling out AI internally should be documenting governance now, not after a mandate lands.

The Decoder↗ source

AI is eroding office hours, study groups, and the trust between faculty and students, MIT report finds

If the university most associated with AI's rise is seeing its own mentoring structures erode, that's a warning for any organization leaning on AI to replace collaborative work rather than support it.

The Decoder↗ source

HackerRank's AI interviewer offers a glimpse into what job interviews could become

Half a million interviews run and Snowflake and Capgemini already testing it. HR and ops leaders hiring at volume should expect candidates to be screened by AI before a human ever sees the resume.

TechCrunch AI↗ source

Anthropic is quietly becoming America's biggest corporate donor ahead of its mega IPO

A philanthropy program that multiplies employee stock donations is unusual optics right before a public listing. Worth noting as one more data point on how AI wealth is being deployed ahead of scrutiny.

The Decoder↗ source

AI glasses face their first major government crackdown

Norway pausing deployment to write permanent rules on recording bystanders is an early marker for how fast wearable AI regulation could move elsewhere. Any ops team piloting smart glasses for warehouse or field work should watch this closely.

Ars Technica AI↗ source
Past issues
The Manifest · 2026-10-06Follow on LinkedIn