In Today’s edition:

1. Anthropic CEO Says AI's Biggest Battle Is Public Trust

2. PMNA TUTORIAL: How to Build an Honest Parent-Status Rollup from Subitem Reality on Your Monday.com Board

3. Asana Uses OpenAI’s Codex to Finish a Five-Year Project in Two Weeks

4. TOOL REVIEW: Crew Tower Helps You Monitor AI Agents Without Losing Focus

5. Exciting Career Opportunities for Product and Project Management Professionals

Reading time: 5 minutes

HEADLINE NEWS

Anthropic Introduces Watermarks to Make AI Content Easier to Trace

Anthropic CEO Dario Amodei argues that the growing backlash against artificial intelligence is not primarily caused by fears about AI itself. Instead, he believes it reflects a much deeper and broader crisis of public trust in governments, corporations, and the technology industry.

According to Amodei, AI companies cannot solve this problem through marketing campaigns or public relations. They must earn trust by delivering meaningful, measurable benefits that improve people's lives.

• Amodei said the current AI backlash is fundamentally a "crisis of trust" rather than a rejection of the technology itself.

• He argued that people increasingly believe corporations will use new technologies in ways that primarily benefit themselves.

• The Anthropic CEO rejected the idea that AI companies can repair their reputations through better messaging or marketing.

• He said the industry must deliver tangible achievements, pointing to breakthroughs such as helping to cure cancer as an example of AI's potential.

• The debate comes amid growing concerns about AI's effects on jobs, privacy, transparency, safety, and corporate accountability

Amodei's comments highlight a growing challenge for the AI industry: technological progress alone is no longer enough to win public support. AI companies must demonstrate that their systems create real-world value while operating transparently and responsibly.

His message is clear: trust cannot be manufactured through advertising. It must be earned through results, accountability, and benefits that people can clearly see and experience. Read More 

ONE MONTH FREE OF MAKE.COM!

If you’ve been looking for a smarter way to automate your workflows, now’s the perfect time to try Make. It’s one of the most powerful tools for connecting apps, streamlining tasks, and building systems that run without you.

We have brought you a fine solution. This gives you full access to advanced features to enable you to automate lead gen, content distribution, or internal operations.

For a limited time, you can get 1 month of the Pro plan completely free using my link below:

PMNA TUTORIAL: How to Build an Honest Parent-Status Rollup from Subitem Reality on Your Monday.com Board

Open any mature Monday board, expand a green parent item, and you'll often find eight subitems done and two open; one of which is the QA gate nothing ships without. The parent row feeds your dashboards, filters, and Thursday sponsor conversation, so when it lies, everything downstream inherits the same confident falsehood. This build watches subitem changes, has Claude read the whole set the way a sharp PM would, and writes an honest judgment back to Monday as both a status colour and a plain-English sentence.

Key Steps and Ideas:

1. Collect your IDs first — column IDs aren't titles, and you'll need the status column's exact labels and indexes before anything else works.

2. Register the webhook on the main board — use change_subitem_column_value, which hands you parentItemId in the payload; registering it on the subitem board gets you nothing.

3. Build the loop guard immediately — drop events whose userId matches your integration, and whitelist only the columns that actually move the judgment.

4. Debounce the burst — queue by parent ID and evaluate once per quiet window, so six Friday-afternoon edits don't become six contradictory judgments.

5. Fetch the full picture — one GraphQL call for the parent plus every subitem; the change event is only a trigger, not the input.

6. Run the judgment call — Claude identifies the load-bearing subitem, weighs remaining work against the deadline, and returns a status constrained to your board's exact labels plus a one-sentence driver.

7. Write both fields back — status colour for machines, long-text summary for humans, in a single change_multiple_column_values mutation.

8. Respect the human override — a "manual status" checkbox that skips the write, so deliberate PM decisions never get stomped.

9. Log and calibrate — track whether items flagged At Risk actually slipped, and trace bad calls back to their inputs.

What You Should Consider:

The information needed to know that release was at risk sat on the board the whole time, one click away, in five subitems anyone could read. It never reached the parent row because the only two mechanisms available are a human who has to remember to look and a rule that can only count, and neither can weigh a QA gate against a README. Put Claude between the subitems and the parent status, constrained to your real labels and guarded against its own echo, and the top row stops being where truth gets rounded off and starts being where it first shows up.

FULL GUIDE HERE

Asana Uses OpenAI’s Codex to Finish a Five-Year Project in Two Weeks

Asana has revealed how it used OpenAI Codex to complete an engineering project that was originally expected to take five years in just two weeks. The company used AI coding agents to replace an outdated testing framework that had become a major obstacle to modernizing its software infrastructure.

The project cost approximately $12,000 in AI model and infrastructure expenses, compared with an estimated $6 million under the previous staffing plan. The experience dramatically changed how Asana evaluates large-scale engineering projects.

• Asana used OpenAI Codex to remove Enzyme, an outdated testing system that was slowing frontend modernization.

• The migration project was originally expected to require five years of engineering work.

• AI agents completed the work in approximately 1.5 weeks spread across two calendar weeks.

• Up to four coding agents worked simultaneously in separate copies of the codebase after receiving a simple five-sentence prompt.

• Human engineers remained in the loop, reviewing progress twice a day and approving every proposed code change.`

The case demonstrates how AI agents are beginning to change the economics of software engineering. Projects once considered too expensive or too time-consuming may now become realistic with AI-assisted development.

However, Asana's CTO emphasized that not every multi-year project will automatically shrink into a two-week task. Instead, the larger lesson is that AI can give engineers more freedom to tackle ambitious projects that previously seemed impossible. Read More 

TOOL REVIEW: Crew Tower Helps You Monitor AI Agents Without Losing Focus

Crew Tower is a macOS utility designed to help users manage multiple AI coding agents without constantly switching between terminal windows. It provides a centralized view of agent activity, progress, context usage, approvals, questions, and completed tasks directly from the Mac’s notch or a floating pill, making it easier to supervise several AI agents while staying focused on the work at hand.

• Multi-Agent Monitoring — Tracks multiple AI coding agents including Claude Code, Codex, Cursor, Gemini CLI, Qwen Code, and opencode, giving users a single place to see which sessions are working, waiting, or finished.

• Centralized Agent Approvals — Lets users approve or deny commands, answer agent questions, and review plans directly from the Crew Tower interface instead of repeatedly returning to individual terminals.

• Live Progress and Context Tracking — Shows session activity, completion status, subagent progress, and context-window usage, including warnings as an agent approaches its available context limit.

• Direct Terminal Navigation — Allows users to jump directly to the exact terminal tab or pane running a particular agent, even when several agents are working within the same project or directory.

• Local-First Operation — Keeps code, prompts, and transcripts on the user's Mac while communicating with agents through local channels, with no account or analytics required; it also supports privacy mode for screen sharing.

Crew Tower is relevant to project managers because it offers a useful model for supervising AI-driven work without needing to monitor every individual process manually. As development teams increasingly use multiple AI agents for coding, testing, research, and other technical tasks, tools that provide centralized visibility into progress, blockers, decisions, and completion can make AI-assisted workflows easier to coordinate and oversee.

TOOL LINK

Exciting Career Opportunities for Product and Project Management Professionals

• Product Owner- Cyber Security IAM @ Wells Fargo

  📍 Charlotte, NC.

• Program Manager - Growth & Practice Partnerships @ Duke Health

  📍 Durham, NC.

• Founding Product Lead @ Atomic

  📍 Charlotte, NC.

• AVP, Technology Product Manager - Sales @ MetLife

  📍 Cary, NC.

• Sr. Product Owner - GM Motorsports @ General Motors (GM)

  📍 Concord, NC.

THAT’S A WRAP