Most companies think AI implementation is a technical problem. Install the right tools, train employees on basic prompts, and results will follow. But here’s the reality: 85% of AI initiatives fail—and the issue isn’t the technology. It’s the human element.

In Episode 2 of Candid Conversations: Marketing and AI Unscripted, Deb Andrews and Brady Lewis tackle the uncomfortable truth about AI transformation: companies are stuck on what they call the “Human to Machine Scale,” unable to move beyond basic adoption because they’ve overlooked change management, strategic thinking, and the psychological barriers holding teams back.

If your organization has dabbled in AI but hasn’t seen transformative results, this conversation offers a roadmap—and a reality check.

Understanding the Human to Machine Scale

The Human to Machine Scale provides a framework for understanding where your organization sits in the AI adoption journey:

• Level 0: All human (pre-generative AI)
• Level 1: Mostly human with some AI assistance
• Level 2: Half human, half machine
• Level 3: Mostly machine with human oversight
• Level 4: Fully automated with AI agents

Most companies? They’re stuck at Level 0 or barely into Level 1. And according to Brady, the biggest obstacle isn’t technical capability—it’s change management and human psychology.

“Too many people, too many companies think that AI is just going to be this magic switch that they can flip on and all of a sudden everything’s going to work,” Brady explains. “That’s not really how it works.”

Why AI Implementation Fails: The Human Factor

The conversation reveals a critical insight: technology readiness has outpaced human readiness. Companies have access to powerful AI tools, but they lack the strategic frameworks, upskilling initiatives, and change management processes to use them effectively.

Brady identifies several ways this manifests:

Lack of Clear Direction: Many organizations have told employees to “learn AI and use it to be more efficient” without providing guidance on which tools to use, how to use them, or what the company’s AI strategy even is.

Hidden Usage and Lost Insights: When companies don’t explicitly give permission and direction for AI use, employees hide their experiments—afraid they’ll get in trouble. This prevents organizations from learning what works, sharing best practices, and building consistent processes.

The Zero-to-Hero Fallacy: Some companies think they can jump from Level 0 to Level 4 overnight. “That’s just a recipe for disaster,” Brady notes. The successful companies are the 15% taking incremental steps, learning as they go.

Deb reinforces this point: “Organizations are not gonna see a step change in efficiency if you have a hundred people using a wide range of tools in various ways.” Without strategic coordination, AI adoption creates chaos rather than productivity gains.

The Roadmap: Moving from Level One to Level Two

So how do companies break through? Brady offers two practical recommendations:

1. Form an AI Council

Before rolling out tools across your organization, create a dedicated group that includes senior leadership and enthusiastic team members from different departments. This council becomes your sounding board, experimentation team, and knowledge-sharing hub.

The key is getting buy-in at the top while empowering people across the organization to experiment and report back—both successes and failures.

2. Start with One Low-Risk, High-Impact Project

Don’t try to “boil the ocean.” Instead, identify a single repetitive task that doesn’t require creative strategic thinking—something like meeting summarization or note organization.

“Start with something that’s not going to crush you if it messes up,” Brady advises. “Something that’s not critical, something that’s not going to clients. Something that’s internal.”

This approach provides several benefits:

• It builds momentum and proves AI can work
• It removes tasks employees don’t enjoy anyway, reducing psychological resistance
• It creates a template for future AI implementations
• It generates confidence in both leadership and staff

The companies succeeding with AI aren’t the ones attempting wholesale transformation overnight. They’re the ones taking deliberate, strategic steps—learning, adjusting, and scaling gradually.

Artisans and Explorers: Staying Irreplaceable in the AI Era

The second major theme from this conversation addresses a fear many professionals face: if AI can write, research, and analyze, what makes me valuable?

Deb introduces a compelling framework: to remain irreplaceable, professionals need to become both Artisans and Explorers.

Be an Artisan: Master Your Craft

“You have to be great at the thing that you do,” Deb emphasizes, “because if you’re not, let’s face it, ChatGPT—other large language models like having a PhD in your pocket—will be able to do technical things.”

In the age of AI, baseline competence isn’t enough. You need to be the LeBron James, the Michael Jordan, the Michael Phelps of your field. Deep expertise allows you to:

• Craft more nuanced, specific prompts that generate better results
• Recognize when AI is hallucinating or providing incorrect information
• Add the creative, strategic thinking that differentiates great work from generic output

Be an Explorer: Gain Unique Experience

Technical mastery alone isn’t sufficient. You also need real-world context that AI simply cannot replicate.

Deb uses Jane Goodall as the perfect example: “No one could replace Jane Goodall because she lived with chimpanzees. She observed their behavior. No one else did that. ChatGPT can’t replace Jane Goodall ever because she was an explorer in her field.”

For marketers, CPAs, engineers, and other professionals, this means:

• Getting face-to-face with clients to understand their challenges
• Observing how your recommendations play out in practice
• Building contextual understanding that only comes from direct experience
• Developing perspectives that are uniquely yours

When you combine artisan-level expertise with explorer-level experience, you create a value proposition AI cannot touch. You’re not competing with the technology—you’re using it to amplify insights only you can provide.

Tool Spotlight: Why Claude AI Deserves Your Attention

The episode wraps with an enthusiastic endorsement of Claude AI, which both Deb and Brady call “the most underrated tool out there.”

While ChatGPT dominates mindshare with roughly 99% market share, Claude offers distinct advantages:

Superior Writing Quality: Brady notes that Claude has consistently produced the most natural, non-robotic writing among all major AI models.

Artifacts Feature: Claude can generate interactive apps, dashboards, games, and visualizations directly within conversations. Upload a dense report, ask for an interactive dashboard, and within minutes you have a functioning tool you can share.

Brady demonstrates this with a real Marketri use case: Deb created a comprehensive website best practices checklist spanning six or seven pages. Brady dropped it into Claude, which generated an organized, interactive checklist with categories—transforming a static document into something teams could actually use in meetings and collaborate around.

“There’s endless usage for Claude,” Deb reflects. “You can do the research, it can put together what you’re going to do… it could be all in one conversation.”

For marketers specifically, Claude’s ability to quickly generate lead magnets, educational tools, and interactive content represents a game-changer. What used to require thousands of dollars and weeks of development time now takes 15 minutes and a well-crafted prompt.

Key Takeaways for Marketing Leaders

This conversation offers several critical insights for anyone leading marketing teams or implementing AI:

Stop Waiting for Perfect Strategy: Form your AI Council and start with one project now. The 15% of companies succeeding aren’t waiting—they’re learning by doing.

Address the Human Element First: Technology without change management, clear communication, and psychological safety creates resistance and chaos. Strategy comes before tools.

Upskill Leadership First: If leaders don’t understand AI capabilities and limitations, they can’t provide strategic direction or identify opportunities. You also protect yourself from vendors who might take advantage of knowledge gaps.

Become an Artisan and Explorer: Don’t let AI do your thinking. Use it to amplify your deep expertise and unique real-world insights.

Experiment with Claude: If you haven’t tried tools beyond ChatGPT, Claude deserves attention—particularly for content creation, research synthesis, and generating interactive elements.

Moving Forward: Your Next Steps

The path from Level 0 or Level 1 to meaningful AI transformation isn’t linear, and it isn’t purely technical. It requires strategic thinking, change management, genuine upskilling, and a willingness to start small while thinking big.

As Brady reminds us: “The people that are trying to do that [wholesale transformation], those are the 85% that we’re seeing failing. It’s the people that are taking it step by step, doing these simple initiatives that actually make a difference that are the ones succeeding.”

Watch previous episodes of Candid Conversations or explore our AI and marketing resources for more insights on navigating this transformation.