Using an AI News Agent to Keep Up With AI - A Complete Guide

Keeping up with AI has become such a challenge that we need AI to solve it. And OpenAI's new Agent may be exactly the right tool for this! Here is why, and how.

The Illusion of Familiar Revolutions

There’s a subtle trick our minds play when faced with technological change: we reach for the most familiar metaphor. In this case, they're not helping.

When we talk about the digital revolution, we often draw comparisons to electricity, to the computer, to the printing press. Each of these was a seismic shift in its own right, reshaping how we live, work, learn, and connect. And yet, with a little distance, each looks almost like a stable, finished thing. The initial chaos of change settled into a steady rhythm of applications opening new ways of consuming, working and living our lives.

But the core revolution itself was stable. Printing, invented half a millennium ago, is now omnipresent in our lives, but the mechanics and possibilities were grasped, understood, and then expanded in a mostly linear fashion. Electricity is still electricity, 100 years later. It’s safer, more distributed, and woven invisibly into everything, but its definition has not changed since Edison. The computer you use today is thousands of times faster and more networked than the one in 1990, but the metaphors still hold: files, folders, desktop, keyboard.

Staying on top of these revolutions, once you learned the basics, was hard but not impossible. The world kept changing, but the ground stopped moving under our feet.

Why AI Breaks the Metaphor

AI is not the printing press or electricity. It’s not a static infrastructure, a finished invention based on a stable definition that merely accelerates and spreads. AI is a living, evolving ecosystem. A constantly morphing, self-upgrading universe of tools and capabilities that seem to leap ahead in unpredictable jumps. What makes this different isn’t just the pace, but the very nature of the change. The ground seems to never stop moving under our feet.

What we’re living through with AI is unprecedented not just in technology, but in the very act of keeping up. The nature of the “revolution” is itself unstable. What’s new isn’t just what AI can do, but that what it can do keeps changing on a weekly, sometimes daily basis. It’s not a challenge of adaptation, but of perpetual reinvention which has become a frequent news topic associated with "AI Fatigue".

In recent months, and as most organizations had barely gotten used to LLMs, GPTs, and copilots, a new paradigm has emerged: AI Agents. And it turns everything upside down again.

It’s Impossible to Keep Up Alone

If you feel like you can’t keep up with AI advancements, you’re not alone. The people building this technology often feel the same and quite frankly - so do I! Here’s why:

1. Constant Feature Emergence: Every month, something truly new drops. Agents, code interpreters, real-time connectors, autonomous research tools, video creators, multi-agent systems, no code platforms and all sorts of copilots for every job function.
2. No Plateau or Deceleration: Unlike other technologies, AI isn’t settling into a mature, slow-evolving infrastructure. If anything, it’s accelerating and each leap changes the landscape of work, possibility, and risk.
3. An Ever Expanding Scope: AI doesn’t just transform software or tech jobs, it's not squarely restricted to knowledge workers either as self-driving cars are increasingly demonstrating. It’s impacting every sector, every role from legal to biotech, education to logistics, creativity to compliance, medicine to driving.
4. A Meta-Learning Imperative: To master AI, you don’t just learn a tool. You have to learn how to keep learning as the tools morph. This can feel exhausting, especially for busy workers, leaders and practitioners.

As Head of AI Products & Innovation at Moderna, this challenge is not theoretical to me. My team’s job is to make sense of AI’s relentless evolution and to make it actionable for our organization. We need a way to not just adapt, but to be on top and lead in the AI age.

Using AI to Keep Up With AI - the "AI News Agent"

Here's the simple truth: only AI can help us keep up with AI!

The Problem

There’s more information than ever, but less clarity. Most AI newsletters are either breathless clickbait or dull recaps. They miss nuance, can’t distinguish hype from real signal, and rarely give actionable advice for business or innovation leaders. Much less customized guidance for your specific role, context or need.

Nobody wants another layer of standard summaries, we want curated intelligence, ranked by relevance to our evolving needs, synthesized for action, and delivered in a way that would stand up to scrutiny among AI Champions, Business Teams, and the C-suite.

The Solution

A concrete example is the “AI News Agent,” an recurring task that uses OpenAI's new Agent capabilities to gather, score, and synthesize the most important news about AI advancements every single week, using a transparent and rigorous method.

You can find the complete prompt here - this link generates a private copy for you, so you can go ahead and play with it. And at this point all you need to do is copy and past the content of the document to see it in action. If you have access to Agent as part of your "Pro" ChatGPT Subscription, that's awesome. If you don't, keep reading! This instruction will also deliver great results with reasoning models such as o3 and o3-pro.

Step 1: Give the Agent a Mission

Define a recurring task for the Agent: In this case, every Friday at noon EST, surface and explain the 15 most consequential AI stories from the week. Stories that matter for innovation, leadership, and business impact. Don’t just collect headlines but rank, explain, and advise with a lens that is specific for my context and needs.

Step 2: Codify What Matters using "Topic Tags"

Not every AI story is created equal. Here are 10 topic tags that reflect my own topics of Interest with AI transformation:

1. AI Usage Levels (insight engine, work orchestrator, life catalyst)
2. Adoption Momentum (usage, satisfaction, business impact)
3. Role Redefinition (future of work, new roles)
4. Cultural Change (data-driven, hybrid intelligence)
5. AI as Teammate (from assistant to creative partner)
6. Systems of Intelligence (agentic workflows, delegated autonomy)
7. Responsible AI (ethics, safety, inspiring use cases)
8. AI for Innovation (breakthroughs, new concepts, trends)
9. Global Impact (climate, pandemics, resilience)
10. AI × Health & Life Sciences (pharma, biotech, life science platforms)

Every story must be tagged with at least three of these, and every tag must appear at least once across the deck. This enforces breadth, depth, and relevance.

Needless to say, you can and should customize the topic tags to meet your own interest and in return create your own bespoke News Agent - on AI or any other topic that you want to constantly remain not only on top, but ahead of.

Step 3: A Scoring Formula for Signal Over Noise

To rank stories, the Agent uses a scoring formula that considers:

• Number of relevant topic tags (depth)
• Source quality (Tier 1: trusted, high-frequency, paywall-free)
• Human productivity gains
• Readiness for economic turbulence
• Recency
• Originality

Only the best stories make the cut—those with a score of 8 or higher, and at least three relevant tags.

Step 4: An Actionable Story Format

For each story, the Agent delivers:

• What: A concise factual summary.
• So What: The implications and possible consequences, using game theory thinking.
• Now What: Forward-looking, actionable advice.

Each entry is JSON-ready, with source, date, tags, and URL so that it is auditable, parsable, and ready for decision-making.

Step 5: Step-Back Prompting To Surface the Week’s Core Themes

Before sharing, the Agent is instructed to “step back” and look across all 15 stories to extract the three most important emerging themes. What patterns are emerging? What new inflection points or systemic changes are visible?

It then raises three questions not just for reflection, but as seeds for further curiosity and deeper inquiry. In our workflow, these can then be further explored using Curiosity GPT, an agent designed to dig deeper on “why” questions and strategic analysis.

Step 6: Publish the Story Deck And Top Five Must-Reads

To make it shareable, the Agent distills the week’s top five articles, each with a one-line blurb and direct link. Something anyone can post to social media, internal Slack, or executive summaries.

Why This Matters And What’s Next

AI isn’t slowing down. In fact, we’re seeing “meta-acceleration” as AI is now building tools to accelerate AI itself. This is exactly what’s happening with the emergence of Agents.

It’s tempting to see this as overwhelming and dystopian. But I see it as a call to action: Use AI to keep up with AI, and stay ahead. It’s not just about reading the news; it’s about living the transformation, building workflows that are agentic, dynamic and adaptive. This is how leaders, researchers, and creative professionals can focus not just on “what’s new,” but on what matters.

Conclusion: The Pace of Change Is The Change, AI Is The Solution.

We’re living in a brand new world, one where the pace of change is itself the biggest change. We can no longer rely on yesterday’s information methods. Instead, we must build living ecosystems of intelligence that include human plus AI, learning together, scanning the horizon and sharing what we find.

The real revolution is not just in what AI can do, but in how we use AI to support our "Why". Let’s use Agents to surface bespoke intelligence, keep each other informed and smart in how we find insight, get things done and thrive!

If you’re interested in seeing how this works, subscribe to the “Power of Why” newsletter for upcoming releases, demos, and hands-on guides. And if you’re building your own AI workflows, I’d love to hear your story, please reach out and share your challenges and accomplishments!