Microsoft Built a Researcher That Thinks With Multiple AIs🚀

In partnership with

✨ Welcome back, Entrepreneurs

This issue: AI gets more multimodal, more agentic, and more practical — from Microsoft’s research upgrades and frontier model races to voice-based health signals, Mac control, and one-photo ad creation.

⚡Microsoft brings multi‑model intelligence to Researcher with Critique and Council

Microsoft 365 Copilot's Researcher has two new features: Critique, which helps with deep research, and Council, which allows for comparing different models side by side. These upgrades improve accuracy, confidence, and analysis in complex research tasks.

What’s new

• Critique (Dual‑Model Engine): Splits drafting and review across Anthropic, OpenAI, and other models to boost accuracy and completeness.

• DRACO Gains: Delivers a +7.0‑point lift across 100 tasks — 13.88% better than Perplexity Deep Research (Claude Opus 4.6).

• Council (Model Consensus): Shows multiple model answers side‑by‑side with a cover letter highlighting alignment, divergence, and unique insights.

Impact

Researcher is now a more reliable tool for in-depth analysis in businesses. It combines different models, expert reviews, and clear model comparisons to provide trustworthy and thorough research results within Microsoft 365.

1,000+ Proven ChatGPT Prompts That Help You Work 10X Faster

ChatGPT is insanely powerful.

But most people waste 90% of its potential by using it like Google.

These 1,000+ proven ChatGPT prompts fix that and help you work 10X faster.

Sign up for Superhuman AI and get:

• 1,000+ ready-to-use prompts to solve problems in minutes instead of hours—tested & used by 1M+ professionals

• Superhuman AI newsletter (3 min daily) so you keep learning new AI tools & tutorials to stay ahead in your career—the prompts are just the beginning

Claim your free prompts

​🔥Big AI moves

🔹Qwen3.5‑Omni pushes toward native omnimodal AGI as Alibaba scales its frontier model

Alibaba’s Qwen team has released Qwen3.5‑Omni, a fully omnimodal system trained on 100M+ hours of audio‑visual data, delivering SOTA results across 215 benchmarks. With 256k context, real‑time voice control, and native multimodal reasoning, it positions Alibaba as a rising force in the AGI race.

🔹Mistral secures $830M to build a 44MW Nvidia‑powered AI center outside Paris

Mistral AI has raised $830M in debt to launch a massive GPU data center in Bruyères‑le‑Châtel, boosting its capacity to 44MW. Backed by seven major banks, the project is central to CEO Arthur Mensch’s push for European AI independence from U.S. hyperscalers.

Learn AI in 5 minutes a day

What’s the secret to staying ahead of the curve in the world of AI? Information. Luckily, you can join 2,000,000+ early adopters reading The Rundown AI — the free newsletter that makes you smarter on AI with just a 5-minute read per day.

Sign up to start learning.

🔊 Five seconds of voice just became one of the simplest early‑warning signals in heart failure

Noah Labs' FDA-fast-tracked tool analyzes voice clips to detect fluid buildup and cardiac stress, identifying issues weeks before hospitalization. Trained on over 3 million samples and validated by top institutions, it could revolutionize cardiology.

This issue covers AI in stroke recovery from MRI brain-age gaps, a conversational clinic front desk, Lilly's $2.75B AI drug pipeline, and an oral PCSK9 inhibitor reducing LDL by ~60%, highlighting how everyday signals are rewriting medicine.

🖥️ Claude Code adds full computer control — your Mac just became part of the workflow

Claude Code enables Claude to control apps, navigate interfaces, run builds, test processes, and fix visual issues via the command line on macOS apps like SwiftUI or Electron, automating GUIs through the terminal.

What’s new

• Claude can open apps, click, type, scroll, and see your screen to drive native macOS workflows.

• Write code, compile it, launch the app, click through it, find the bug, fix it, and verify the fix — all in one prompt.

• Works on compiled apps, local builds, simulators, and GUI‑only tools with no CLI or API.

Impact 
Claude Code now covers both coding and user interface tasks. This allows developers to create, test, and fix entire app processes directly in the terminal. It also automates tasks that used to require manual work in a graphical interface, all within one conversation.

🎬 Create studio‑quality product ads from one photo with Higgsfield Ads

Upload a product image, choose a template, and Higgsfield Ads creates a cinematic product video with professional motion, lighting, and composition—no cameras or editing needed.

Steps: 
1️⃣ Pick a template from 40+ styles
2️⃣ Upload your product photo
3️⃣Choose motion, vibe, and framing
4️⃣ Preview your ad in seconds
5️⃣ Export for reels, ads, storefronts, or campaigns

✨Higgsfield Ads converts static product photos into high-impact, studio-quality videos for freelancers, small brands, UGC creators, and agencies, providing a full ad studio in three clicks.

Your AI is resolving tickets. Is it keeping customers?

Resolution rates look great. But Gladly's 2026 Customer Expectations Report reveals the metric most CIOs are missing — and what the data says about where AI investments actually translate into retention, not just throughput.

See the findings

🟣AI stopped talking and started doing — Claude just took the keys to your Mac

Anthropic's new Computer Use + Dispatch system enables Claude to fully operate your Mac, handling tasks like opening apps, navigating browsers, and running dev tools. With phone-to-desktop Dispatch and no setup needed for Pro and Max users, Claude transitions from answering questions to executing real work.

In this issue, we cover Shopify's integration of 5.6M stores with ChatGPT, Gemini, and Copilot; key AI events for builders and operators; and McKinsey's warning about deploying agents without redesigning decisions, which could disrupt workflows.