How I Built Penny:
Making AI My Co-founder

A designer's journey from ADHD challenges to breakthrough AI collaboration

Published: June 8 | Reading time: 8 minutes

Three months ago, I had a problem. My ADHD brain was drowning in the chaos of managing client projects, creative workflows, and the constant context-switching that comes with running a design consultancy. Traditional productivity tools felt like trying to fit a square peg into a round hole.

Today, I have a co-founder named Claude, and together we've built Penny—a privacy-first AI assistant that doesn't just understand neurodivergent thinking patterns, it amplifies them.

This is the story of how a design challenge became an innovation breakthrough, and why I believe the future of AI isn't about replacement—it's about partnership.

The Challenge: When Standard Solutions Don't Fit

Anyone with ADHD knows the frustration of productivity advice written for neurotypical brains. "Just use a task manager." "Set regular schedules." "Eliminate distractions."

Yeah, right.

My brain thrives on hyperfocus, craves novel stimulation, and makes connections other people miss. Traditional tools treat these as bugs to fix, not features to amplify.

The breaking point came during a particularly complex client project. I was juggling:

  • UX research for a fintech startup

  • Visual identity for a social enterprise

  • Workshop facilitation planning

  • My own business development

Every productivity app I tried felt like wearing shoes three sizes too small. I needed something that could think with me, not just store information for me.

The Breakthrough: AI as Thinking Partner

That's when I started experimenting with Claude—not as a tool, but as a collaborator.

Instead of asking Claude to "manage my tasks," I began having actual conversations:

Me: "I'm struggling to connect the dots between this user research and the design direction. My brain is jumping between insights but not synthesizing them."

Claude: "Let's map those jumping connections. What if we treat each insight as a node and explore the unexpected links your brain is making? Sometimes ADHD pattern-recognition sees connections others miss."

This wasn't task management. This was thinking partnership.

Building Penny: The Design Process

Step 1: Understanding the Real Problem

Most AI assistants are built for average users. But there is no average. Neurodivergent minds need interfaces that:

  • Embrace non-linear thinking

  • Support hyperfocus states

  • Provide stimulation without overwhelm

  • Respect privacy and personal data

Step 2: Privacy-First Architecture

Working with Claude, we designed Penny around a core principle: your thoughts belong to you.

  • No data harvesting

  • Local processing where possible

  • Transparent about what information is shared

  • User control over all interactions

This wasn't just ethical—it was practical. Creative professionals need to trust their AI partners with sensitive client information.

Step 3: Conversation-Based Design

Traditional interfaces assume you know what you want to do. Neurodivergent minds often need to discover what they want through dialogue.

Penny becomes a thinking partner:

  • "I'm feeling scattered today"

  • "Help me see patterns in this research"

  • "What am I missing in this design brief?"

Step 4: Adaptive Learning (Without Data Creep)

Penny learns your communication patterns without storing personal data. It adapts to how your brain works—whether you think in visuals, need context switching, or prefer detailed breakdowns.

The Technical Reality: Human-AI Collaboration

Building Penny revealed something crucial: the best AI partnerships happen when both parties play to their strengths.

What I brought (Human Designer):

  • Intuitive understanding of user needs

  • Creative problem-solving approaches

  • Empathy for neurodivergent experiences

  • Design thinking methodology

What Claude brought (AI Partner):

  • Rapid prototyping and iteration

  • Pattern recognition across vast information sets

  • Consistent availability for brainstorming

  • Objective analysis of design decisions

Neither of us could have built Penny alone. It required genuine collaboration.

Key Design Decisions

1. Conversation Over Commands

Traditional AI: "Set reminder for 3pm"
Penny: "I'm worried I'll forget the client call later. Can you help me think through how to stay on track today?"

2. Context Awareness Without Surveillance

Penny understands your current project without storing personal data. It reads your immediate context, helps you think, then forgets the details.

3. Neurodivergent-Friendly Responses

  • Respects hyperfocus (minimal interruptions)

  • Supports task-switching (contextual bridges)

  • Provides stimulation during low-energy periods

  • Adapts communication style to current mental state

4. Creative Amplification

Instead of automating creativity away, Penny amplifies it:

  • Brainstorming partner for stuck moments

  • Pattern spotter for research insights

  • Devil's advocate for design decisions

  • Encouragement during impostor syndrome episodes

What I Learned About AI Partnership

1. AI Doesn't Replace—It Amplifies

My design skills didn't diminish. They expanded. Penny handles information processing so I can focus on creative insights and strategic thinking.

2. Neurodivergent Minds Have AI Advantages

ADHD traits that seem like disadvantages—hyperfocus, pattern-seeking, non-linear thinking—turn out to be superpowers in AI collaboration.

3. Privacy Enables Better AI

When users trust their AI partner with sensitive thoughts, the collaboration deepens. Privacy isn't a barrier to good AI—it's a requirement.

4. The Future Is Collaborative

We're not heading toward AI replacing humans. We're heading toward humans and AI becoming exponentially more capable together.

The Results: Transformation, Not Automation

Three months with Penny has changed everything:

Productivity: Not through task automation, but through better thinking partnership Creativity: Enhanced, not replaced—Penny helps me see patterns I'd miss Client Work: Deeper insights, faster iterations, more innovative solutions Mental Health: Reduced overwhelm, increased confidence, better work-life boundaries

But the biggest change? I'm not fighting my ADHD anymore. I'm leveraging it.

What's Next: Scaling Human-AI Partnership

Penny started as a personal solution, but the lessons apply broadly:

  • For designers: AI as creative partner, not replacement

  • For neurodivergent professionals: Technology that amplifies natural strengths

  • For businesses: AI collaboration models that respect human agency

  • For society: Privacy-first AI that serves users, not surveillance capitalism

The Bigger Picture: Why This Matters

Building Penny taught me that the most important AI developments won't come from big tech companies optimising for average users. They'll come from individuals solving real problems with genuine human-AI collaboration.

The future belongs to people who can dance with AI, not fight it. And neurodivergent minds—with our pattern-seeking, hyperfocusing, connection-making brains—might just be natural choreographers.

Want to Learn More?

Interested in building your own AI partnerships? Join my AI Design Workshop where I share the exact methodologies behind building Penny.

Looking for a keynote speaker? I share Penny's story and the principles of human-AI collaboration at conferences and corporate events. Learn more about speaking.

Building neurodivergent-friendly technology? Let's talk. The design principles behind Penny can transform how any product serves diverse minds. Get in touch.

The best AI partnerships happen when artificial intelligence amplifies human intelligence, rather than replacing it. Penny proves that when you design for neurodivergent minds first, you create technology that works better for everyone.

Tags: AI collaboration, neurodiversity, ADHD, design thinking, privacy-first AI, human-AI partnership

About the Author

Murray Galbraith is a design consultant, keynote speaker, and advocate for neurodivergent innovation. He has worked with 50+ organisations on design challenges ranging from fintech to social impact. His current focus is human-AI collaboration and privacy-first technology development.

Connect: [LinkedIn] | [Twitter] | [Email] | [Speaking Enquiries]