Real Conversations, Real Growth

Five episodes in, our discussions have evolved into something more profound than just tech or process. Each one has become a snapshot of how we’re learning, teaching, and experimenting with ideas together. This episode continues that thread—an honest, curious exploration of how I use AI, reflection, and collaboration to grow as a creator and thinker.

🎥 Watch the full episode of The Blank Page Podcast: Episode 5 to hear the complete conversation—from AI prompts and problem statements to the spiral of data, insight, and impact.

The AIR Experiment

Matthew kicked things off by mentioning our latest AIR meeting (AI Roundtable at Improving), where a teammate presented his latest AI experiments. He’s been training models using Python and Claude to solve math problems and perform reasoning tasks—starting from scratch and narrating his process in real time. It inspired us to revisit how we learn and teach with AI.

We discussed how each of us approaches these projects. I’ve been documenting my own AI-first experiments as videos, blog-style writeups, and full-length walkthroughs—letting colleagues “pick their poison.” Matthew suggested adding repos so others can follow along. I pushed back slightly: rather than copying prompts, people should learn how to think through them. “Don’t give the fish—teach how to fish.”

That insight connected naturally to how I communicate intent to AI tools. Voice dictation came up as a way to show that prompts are really just spoken thoughts—not code or syntax. As I said, “I wasn’t describing a feature; I was describing a problem statement. The prompt was whatever came out of my mouth as I was explaining the problem.”

Planning, Auditing, and AI Feedback Loops

From there, Matthew shared his method for developing plans with AI—creating a plan, then spinning up a second instance to audit it, like bringing in a fresh developer for a peer review. He treats each AI session as a separate collaborator, offering a second opinion. I shared how Cursor’s new plan mode now auto-generates clarifying questions, creating a similar back-and-forth that refines understanding rather than code.

Both of us agreed: the goal isn’t just writing code faster—it’s thinking better.

The Spiral: From Data to Impact

I then unveiled a framework I’ve been developing—the Data to Impact Spiral, as I ponder this quote:

We are drowning in information and starving for wisdom. – Tony Robbins

Here’s how I mapped it:

• Data becomes Information when structured.
• Information becomes Knowledge when understood.
• Knowledge becomes Insight when applied.
• Insights accumulate into Wisdom.
• Wisdom only matters when it creates Impact.

Then the spiral continues—impact leads us back to collecting new data, better information, deeper knowledge.

Matthew visualized it as a continuous forward spiral—not a loop but a progression powered by curiosity and action (which is precisely how I visualize it, too!). We explored how this applies to IT itself: perhaps “Information Technology” should evolve into Impact Technology.

Living Wisdom

The conversation deepened into what wisdom means. Is wisdom without application still wisdom? I argued that lived experience completes the cycle: “If you haven’t done anything with it, it’s not wisdom—it’s just knowledge.”

We explored how culture and language shape understanding. The English word insight has no perfect Portuguese equivalent, so Brazilians often use the English term. This led to a larger reflection on language accessibility, the limits of translation, and how communication gaps affect how people use AI.

Language, Accessibility, and AI

Matthew expanded this to accessibility in general—how not everyone, even native speakers, has the same command of language to express their intent clearly to AI. He connected it to his recent work on web accessibility and screen readers, showing how thoughtful design can open new worlds for people.

Together, we noted that clear intent and precise language lead to better AI outcomes.

From Data to Impact in Practice

The spiral came to life in a real-world example: I described how produce warehouse managers think in data terms—inventory, temperature, and shelf life—but what they really seek is impact: fewer losses, better margins, less waste. “If it ain’t selling, it’s smelling.” That visceral insight grounds the abstract spiral in something real.

I even used NotebookLM to generate user stories based on this produce scenario—covering business, user, customer, societal, and alignment value. The result: richer conversations that start with human pain points, not technical specs.

Pre-Mortems, Personas, and Better Conversations

As the episode wound down, we discussed using AI for pre-mortems (instead of post-mortems)—asking what could go wrong before starting a project. Combined with historical company data, these tools could forecast risks and refine decisions. Personas and AI debates (like NotebookLM’s “Critique” and “Debate” features) become a way to simulate diverse perspectives before real problems arise.

Rapid AI Prototyping: The 30-Second Wow

To end on something concrete, I shared a story: during a casual online conversation about scheduling Tech Fridays (a recurring internal presentation at Improving), I explained the problem to AI Studio—and within 15 minutes had a working app. I call it the 30-second wow: a demo that instantly captures attention by showing what’s possible when you articulate problems clearly.

The old reflex was to create a spreadsheet. Now we can just articulate the problem and let the tool create something real.

Reflection and Wrap-Up

We closed the episode on a human note—a reminder to slow down, have dinner, and be present. Because in the end, the real impact isn’t just in the technology; it’s in how it shapes our days, our habits, and our connections.

This entry was posted on November 7, 2025, 7:16 am and is filed under The Blank Page Podcast. You can follow any responses to this entry through RSS 2.0. You can leave a response, or trackback from your own site.