The Genius That Never Ships: Why Action Bias Matters More Than Brilliant Ideas

Last night, I was watching Watson. You know, the TV show where Dr. John Watson operates in today’s world, sans his famous detective companion. One scene stopped me, and I even took a picture.

Watson is briefing his team about a breakthrough in genome editing. The science is remarkable. They can alter individual genomes to cure illnesses. Revolutionary stuff.

But then Watson drops something along these lines: it’s all irrelevant if we can’t actually use it. Legal restrictions, regulatory barriers, practical limitations… they make genius theoretical rather than transformative.

The science exists. The application doesn’t.
Therefore, it might as well not exist.

That line hit differently because I’d just spent two days at the Portugal Digital Summit here in Lisbon.

The event brought together 147 speakers discussing how artificial intelligence is transforming our economy and society. One conversation in particular has been rattling around my brain since. 🤔

The CEO of Deloitte Portugal, António Lagartixo, said something that most people probably glossed over.

He said that in about eight years, his business will be radically different or even dead. 😬
Not incrementally better. Not slightly evolved. Radically different.

What Deloitte does today, the entire business model, the way they provide value, will be dead. Gone. Replaced by something we’re only beginning to understand.

The Uncomfortable Truth About Change

Here’s what connects these two moments: action bias.

If we don’t take action and make things happen, things will happen to us instead. Watson’s genome editing breakthrough becomes a footnote in a research paper. Deloitte’s traditional consulting model becomes a casualty of competitors who moved faster.

You can have all the genius in the world. You can understand AI better than anyone in your sector. You can see the future clearly.

But if you can’t apply that knowledge, if you can’t execute, you become irrelevant. Your company becomes irrelevant.

PwC’s research confirms this reality: in 2025, company leaders no longer have the luxury of addressing AI governance inconsistently or in pockets of the business, as AI becomes intrinsic to operations and market offerings. The experimental phase is over. The shift is from experimentation to execution, offering businesses new opportunities to drive meaningful change through advances in AI, shifts in workplace dynamics, and consolidation in the technology market.

I’ve heard this argument thousands of times at conferences. “We need to innovate.” “We must embrace digital transformation.” “AI is the future.” The audience nods. They take notes. They go back to their offices and do… business as usual.

Because change hurts. Change is uncomfortable. Change means admitting that what worked yesterday won’t work tomorrow.

Why Most People Stay Frozen

Let me be blunt: most organisations prefer the comfort of slow decline over the discomfort of rapid adaptation.

They’d rather optimise a dying business model than rebuild for a new reality. It’s not stupidity. It’s psychology. Big changes cause big waves, making it tough for individuals to stay afloat, whilst small steps tap into something deeply resonant with the human spirit.

Research shows that 70% of transformation initiatives fail, not because ambition is lacking, but because of the pace and scale of change. Large-scale change overwhelms teams, systems, and budgets before delivering results. We announce grand five-year digital transformation programmes, hire consultants, build elaborate roadmaps—and then watch them collapse under their own weight.

The irony?
The organisations that succeed aren’t necessarily making BIGGER bets. As other speakers have shared on stage, they’re making MORE bets.

The Power of Compounding Small Changes

Here’s what actually works: forget the moonshot. Focus on the 1% improvement.

The concept is simple: if you improve by just 1% daily in any area of work, those small gains compound over time, leading to significant improvement and growth. Mathematically, you’re about 37 times better by year’s end.

That sounds too simple to be true. But look at the evidence.

The British cycling team provides a compelling case study. In the early 2000s, coach Dave Brailsford leveraged the idea of improving every aspect of performance by just 1%, resulting in the team winning the Tour de France six times and dominating Olympic victories, winning 70% of gold medals between 2008 and 2012.

It’s not that they rebuild cycling from scratch. They found 100 small things to improve and improved them consistently. Better bike seats. Optimal tyre pressure. Hand-washing protocols to reduce illness. Sleep optimisation. Nutrition timing. Each change was tiny. The cumulative effect was transformative.

Your business needs the same approach.

From Theory to Practice: What to Actually Do

Right. Enough philosophy. Let’s talk about practical action bias in AI automation.

Start With 100 Small Projects, Not One Big One

Gartner research suggests AI is heading for the trough of disillusionment, with Deloitte’s State of Generative AI in the Enterprise reporting that nearly 70% of organisations have moved 30% or fewer of their gen AI experiments into production.

Why? Because they’re treating AI like a massive infrastructure project rather than a portfolio of experiments.

Here’s your new strategy: launch 100 small AI projects this quarter instead of one enterprise-wide transformation.

Small looks like:

• Automating your meeting notes with AI transcription.
• Using AI to draft first versions of routine emails.
• Implementing chatbots for your three most common customer questions.
• Automating invoice data extraction for your accounts team.
• Using AI to screen CVs for your next hire.

Notice what these have in common? Each takes days to implement, not months. Each costs hundreds of pounds, not millions. Each can fail without sinking your business.

The incremental approach involves implementing digital transformation gradually, in smaller, manageable phases, allowing for continuous improvement and adjustment over time.
You test. You learn. You scale what works. You kill what doesn’t.

Measure Everything (But Keep It Simple)

For each small project, track three metrics:

1. Time saved (in hours per week)
2. Error reduction (percentage decrease in mistakes)
3. User satisfaction (simple 1-10 rating from the people using it)

That’s it. No complex ROI models. No six-month analysis paralysis. If it saves time, reduces errors, and people like using it, scale it up. If not, move on.

BMW’s Innovation Radar exemplifies this approach, monitoring more than 100 projects in real-time and allocating resources dynamically based on performance, potential, and shifting market conditions. This strategy allows companies to nurture both incremental improvements and game-changing innovations.

Create a Two-Speed Innovation Engine

Not everything should move at the same pace.
You need:

Fast track (days to weeks):

• Process automation that touches existing systems.
• AI tools that enhance current workflows.
• Chatbots and simple customer service automation.
• Document processing and data extraction.
• Email and content generation tools.

Slow track (months to quarters):

• Custom AI models for your specific domain.
• Integration with core business systems.
• Industry-specific AI applications.
• Comprehensive data strategy and governance.
• Culture and training programmes.

The mistake organisations make is treating everything like a slow-track project.

Instead of speculative pilots or gimmicky features, businesses should focus on proven applications that offer real value. For example, think personalised customer experiences, predictive maintenance, or enhanced decision-making.

Run 20 fast-track projects for every slow-track one. Build momentum. Show value. Create believers in your organisation who’ve seen AI actually work.

Build Your Personal AI Stack Today

Don’t wait for your IT department. Don’t wait for approval.
Start building your personal productivity stack right now with tools that require zero IT involvement:

For content creation:

• ChatGPT or Claude for drafting, editing, and brainstorming.
• Grammarly for writing enhancement.
• Veo3 for idea visualisation.

For research and analysis:

• Perplexity for research summaries.
• Claude or ChatGPT with web search for deep dives.
• NotebookLM for document analysis.

For productivity:

• Otter, Fathom or Granola for meeting transcription.
• Make with AI for workflow automation.
• Notion AI for note organisation.

Track your time for one week without these tools. Then track one week with them. Calculate the hours saved. That’s your personal ROI that you can demonstrate to leadership.

The 30-Day Challenge

Here’s what you can do in the next 30 days to shift from talk to action:

Week 1: Audit and identify

• List your 10 most time-consuming repetitive tasks.
• Identify which could be automated with existing AI tools.
• Choose three to test.

Week 2: Test and measure

• Implement the three tools.
• Track time saved daily.
• Note friction points.

Week 3: Refine and expand

• Fix what’s not working smoothly.
• Add two more automation projects.
• Document your process.

Week 4: Scale and share

• Calculate total time saved.
• Share results with your team.
• Identify next five projects to tackle.

Incremental innovation contributes to sustainable growth and long-term market relevance through cost-effectiveness, as the financial burden is minimised when optimising existing resources rather than implementing radical transformations.

The Compound Effect in Your Career

This isn’t just about your organisation.
It has implications for your professional survival.

The World Economic Forum reports that 50% of employees will need reskilling by 2025, yet 22% of employees say they’ve received little to no support. Your employer might not save you. You need to save yourself.

Technological advancements, including artificial intelligence, are accelerating the pace of change, and investing in people via digital skills training should be a crucial element of how professionals drive the change process.

Every week, dedicate three hours to:

• Learning one new AI tool.
• Automating one manual task.
• Sharing what you learned with someone else.

That’s 156 hours per year. 156 hours of compounded AI expertise.
In 12 months, you’ll be the person everyone comes to for AI guidance.
In 24 months, you’ll be irreplaceable.

The Choice You’re Actually Making

Let me close with something uncomfortable: inaction is a decision.

When you postpone that automation project, you’re deciding to spend another month doing manual work. When you wait for the “perfect” AI strategy, you’re deciding to let competitors learn faster. When you convince yourself you’ll start “next quarter,” you’re deciding that comfort matters more than survival.
Watson’s genome editing was brilliant. But brilliance locked in a lab helps no one.

Deloitte’s CEO sees the future clearly.
But vision without execution is hallucination.

Because AI offers such transformative potential for new operational and business models, those that pull ahead of the pack (whether AI native companies or established companies that reinvent themselves quickly) will likely stay there, with the growing gap between AI leaders and laggards extending to economies as well.

You don’t need permission to start.
You don’t need a massive budget.
You don’t need executive buy-in for your first 10 experiments.

You need action bias.

Start today with one task.
One tool. One hour.

Then compound it.

Because in eight years, your industry will look nothing like it does today.
The only question is whether you’ll be leading that change or scrambling to catch up.

The genius that never ships is just an expensive decoration.
The average solution that’s deployed tomorrow beats the perfect solution that’s never deployed.

Ship it. Measure it. Improve it. Repeat.

That’s how you win in the age of AI.

What’s the one AI automation you’re going to implement this week?
Drop me a message since I’m genuinely curious what you’ll tackle first.