TAG Weekly Roundup #3

Editor's Note

We're witnessing a shift from, explicit programming to "delegating to intelligence"—where systems adapt rather than just execute. 

From Cosmos Institute's philosopher-builders, to agentic browsers that think for themselves, the question isn't whether AI will reshape work, but whether we're building systems that contemplate and enhance human agency, or… if we’re creating sophisticated dependencies that simply execute faster.

📻News

The Product Compass Bridges AI and Business Analysis

Paweł Huryn's #1 AI and PM newsletter on Substack offers actionable insights that cross traditional role boundaries.

With hundreds of thousands of subscribers and comprehensive guides on AI product management, it's invaluable for Business Analysts working with AI, or interested in product management crossover. Read more

Cosmos Institute Cultivates Philosopher-Builders

Cosmos Institute are training technologists to fuse deep inquiry with technical skill. Their focus on human autonomy, decentralisation, and truth-seeking directly challenges the trend toward "delegating to intelligence". Read more

🧰Tools

Comet Browser - Perplexity’s Agentic Experiment

Currently exclusive to Perplexity Max subscribers, Comet transforms browsing into intelligent task execution. The browser handles restaurant bookings, meeting scheduling, and research compilation whilst maintaining conversation context. 

Early reviews highlight genuine "wow moments" but also hallucination issues! Worth joining the waitlist to test.

Slack Gets Seriously Smarter

New AI tools now summarise channels and automate action items across connected applications. Enterprise search spans Salesforce, Microsoft Teams, Google Drive, and Confluence from within Slack. 

Rather than programming specific workflows, Slack's AI discovers and surfaces relevant information contextually. The integration overhead remains significant, but the productivity gains appear genuine.

📽️Video

MCP, ACA and the Beginning of the End of Explicit Programming

The video [9 minutes] explores how Model Context Protocol (MCP) was the first crack in traditional deterministic programming i.e. instead of explicitly determining how AI should use tools, MCP lets us describe capabilities and allows AI to figure it out! 

Google's A2A takes this further - agents now discover each other, understand capabilities, and negotiate collaboration autonomously.

The presenter (Nate), walks through concrete examples, from sales operations to distributed system challenges, showing how we're moving from explicit programming to "delegating to intelligence." 

Did you Know? Business Analysts Show Remarkable Gender Balance: 46.1% of all Business Analysts in the US are women and 53.9% are men, achieving near-parity in a field where many tech roles show significant gender disparities. This balance suggests the analytical and collaborative nature of BA work attracts diverse talent more effectively than traditional technical roles. Source: Zippia Demographics Research

Till next week.