Instagram | Sponsor this newsletter | Let’s Connect!

The AI waves just keep getting bigger—and this week, we’re standing at the edge of something massive.

Microsoft is quietly redefining productivity with agents that reason like humans, OpenAI’s image tools are pushing the boundaries of visual realism, and major tech expansion plans might be stalling—not from lack of demand, but geopolitics.

Also: a sneak peek at Copilot’s new powers, the rise (and fall) of top image generators, and why enterprise AI might be the stealthy MVP of 2025.

AI Agents That Think Like Analysts, Researchers & Teammates
Microsoft has just quietly shipped some of the most powerful workplace AI tools we’ve seen yet—and they live right inside Office.

The new agents—Copilot Researcher and Copilot Analyst—don’t just follow prompts, they reason through your company’s documents, dashboards, and data to generate useful insights like a real coworker.

Imagine dragging two gnarly Excel sheets into Excel and just typing, “Find key insights about revenue per region.” And boom—it’s done, with charts and a full analysis ready to present.

🧠 Prediction: Enterprise users will unknowingly adopt AI through daily workflows. Copilot isn’t an add-on—it’s shaping up to be the native AI operating system for work.

🔗 Read about Copilot’s latest updates

New Duties Could Stall AI Infrastructure Expansion in the U.S.
A new tariff policy from former President Donald Trump could put a major brake on America’s AI growth engine.

According to Reuters, Trump's reciprocal tariffs—especially on high-tech components from China, Taiwan, and South Korea—are likely to increase the cost of key infrastructure for AI and cloud services.

With so much AI development dependent on data center scale and GPU access, these tariffs could slow the rollout of new AI services, increase API pricing, or delay training of large models.

💥 Takeaway: Infrastructure is the invisible foundation of AI—and U.S. protectionism might cause ripple effects throughout the entire ecosystem.

🔗 Read the full story on Reuters

Reeve Image, GPT-4o, and Recraft V3 Redefine Photorealism
In just one week, the leaderboard for top image generators shifted twice.

First, Reeve Image Half Moon topped the Leaderboard Arena—and then, just days later, OpenAI’s GPT-4o and Recraft V3 took the crown.

These tools don’t just create "nice art." They handle product photography, ad creatives, and even perfect text rendering:

These models are starting to replace not just creatives, but entire workflows. Need a Paris shoot for your clothing brand? You can do it in 10 seconds—with AI.

🖼️ Takeaway: These tools are unlocking new creative workflows in fashion, advertising, editorial, and e-commerce. Visual storytelling just got fully democratized.

🔗 See GPT-4o’s capabilities | Try Recraft | Track the leaderboard

Want to turn a messy spreadsheet into clean charts and insights—without writing a single formula?

Here’s how:

💼 Why it’s powerful:

💡 Pro tip: Ask for a presentation summary after the insights—it’ll write your meeting talking points too.

🔗 Try Copilot in Excel

🕹️ NVIDIA introduces Project G-Assist, a real-time AI assistant that helps gamers optimize settings, performance, and strategy—directly inside their games. Read more

🧠 Anthropic breaks down how language models think, offering a framework for tracing internal reasoning steps in large models—potentially paving the way for better alignment and transparency. Read more

🧱 Hugging Face showcases TripoSG, a model that turns single 2D images into stunning 3D assets—great for games, design, or AR workflows. Read more

🧑‍💻 Amazon unveils Nova, a developer SDK that lets you build web-based AI agents capable of completing tasks in-browser—like shopping, summarizing, and form-filling. Read more

Until our next AI rendezvous,

Anthony | Founder of Uncover AI