When clarity meets AI, anyone can turn insight into a system.
getty
A few days ago, I boarded a flight to Zurich, opened Claude Code for the first time, and six hours later landed with a fully functional AI system that does what used to take me a team and several hours: scan research, synthesize insights, score them for relevance, compile notes from all my calls, and email me a strategic brief every week, complete with clickable sources. md.
I’m not a software engineer. I’ve never taken a computer science class. Until recently, I couldn’t tell you Python from JavaScript. But in the time it takes to watch two movies, I created what my team now calls the OA Agentic Wisdom Engine, a system that automates the discovery and synthesis work that used to consume hours of my week. That experience changed how I think about the future of work, expertise, and who gets to build.
The Experiment
For years at Open Assembly and Harvard Business School, I’ve written about how AI would transform work and expertise. I’ve advised Fortune 500 companies on AI adoption and co-authored articles in this publication on systematically implementing AI. But it’s one thing to study disruption—and another to experience it.
I wanted to know: Could someone with zero coding experience build something useful with these new AI tools? Not just a chatbot or a form-filler, but a real system: multi-step, automated, and actually useful? The answer, I discovered, is an unequivocal yes. And the implications are staggering.
What I Built
The system mirrors the DIKW framework: Data, Information, Knowledge, Wisdom, that’s long guided my advisory work. It works in four stages: Sense, Interpret, Synthesize, and Anticipate, each building toward actionable intelligence.
- In Sense, it scans key sources, HBR, HBS, McKinsey, WEF, Gartner, Deloitte, and transcripts from my weekly public calls, for signals tied to enterprise AI, workforce transformation, and human-in-the-loop governance.
- In Interpret, Claude analyzes each source, extracts insights, and scores each one based on relevance, novelty, actionability, and credibility. Anything that scores above an 8 out of 10 gets flagged for my review.
- In Synthesize, it generates weekly briefs that highlight strategic patterns, offer concrete recommendations, and include clickable links to original sources.
- In Anticipate, which I’m still developing, the system will connect to my decades of archived work in Google Drive to create a contextual knowledge base that grounds every insight in my own thinking.
Everything writes to a Google Sheet dashboard, runs through approval workflows, schedules daily ingestion, and delivers finished briefs to my team every Monday morning.
How It Happened
I used Claude Code, Anthropic’s command-line tool. But the experience didn’t feel technical, it felt conversational.
I started by describing what I wanted: “I’d like to build a wisdom engine.” I uploaded a rough architecture sketch. Claude reviewed it, offered feedback, asked clarifying questions, and suggested a place to begin.
Then it started coding. It wrote a function, tested it, hit an error, fixed it, and explained each step in plain English. When I hit my OpenAI API limit mid-project, Claude pivoted to Anthropic’s own API and rewrote the integration on the fly.
When I asked to add email functionality, it wrote the module, walked me through setting up a Gmail App Password, configured SMTP, and sent a test message. When I said, “Let’s send it to [email protected] and [email protected],” it did within seconds.
The terminal log reads more like a dialogue than a programming session. I wasn’t writing code. I was describing what I needed. Claude did the rest.
What This Means for Leaders
1. The technical barrier is gone.
The line between technical and non-technical is dissolving. I built a production-grade system on a flight. That doesn’t mean engineers are obsolete, but it does mean domain experts can now turn ideas into tools. The real constraint isn’t code. It’s clarity.
2. Human judgment is more valuable, just in new ways.
My system has human-in-the-loop governance built in. AI handles discovery and synthesis. I make the judgment calls. This kind of partnership, AI for scale, humans for meaning, is what will define high-functioning teams going forward.
3. Culture is the differentiator.
If a 60-year-old executive can build a working system in an afternoon, imagine what your employees could create with the right tools and support. The advantage isn’t in the AI itself. It’s in the culture that unleashes people to use it.
The Explorer’s Mindset
I’ve climbed mountains, surfed big waves, and skied steep lines. But opening a command line? That was a different kind of fear, the fear of looking foolish in a domain where I had no credentials or training. That fear was misplaced. The tools have changed. The gatekeepers are gone. The new question isn’t “Can you code?” It’s “Can you think clearly about what you want to build?”
This is what I call the Explorer’s Mindset, the willingness to enter unfamiliar territory and treat uncertainty as opportunity, not threat. That mindset has driven every pivot I’ve made, from publishing to advertising to open talent and now, AI.
The difference now? The cost of exploration is near zero. I didn’t spend months learning Python. I didn’t hire developers. I didn’t wait for IT to approve my experiment. I just started talking to an AI and it helped me build what I imagined.
The Real Takeaway
The real story isn’t the system. It’s what it represents: a shift in who gets to build. For decades, software was the domain of specialists. Today, anyone with insight and a clear idea can create systems that scale their thinking.
In Open Talent, I wrote: the war for talent is over, talent won.
Now, I’d add: the war for technical capability is over, too. The tools have won. The only question is whether you’ll use them.
Six hours. One flight. Zero coding experience. One working AI system.