Conversation on AI, Trust, and the “Hollowing Out” of Professional Services 

Marketing Hype vs. Real Capability 

They start by talking about the recent excitement and short pause around Anthropic’s new AI releases. Nathan is skeptical that the branding, with names inspired by Greek mythology, makes the models seem more powerful than they are. He thinks clever names can exaggerate a model’s abilities. Charles disagrees a bit, saying that even with the marketing, he’s seen impressive results, like 3D design work and fully animated websites and games that would usually take a team weeks to make. 

Where they land is an important nuance for anyone using AI tools in their own business: 

• Charles points out that, with image and video generation, most people can’t tell the difference between AI-generated content and human-made content. 
• The main difference comes from the prompt, not the platform. The quality of the result depends more on how well someone uses the tool than on which model they choose. 
• Big AI labs are now focusing on the toughest problems, such as gaming graphics today and, possibly, medical imaging or accounting tools for small businesses in the future. 

We’ve utilised AI to help us create artwork mock-ups for clients. While we have an in-house graphic designer ready, having AI gives our account managers an option (if the client urgently needs the mock-up). 

Will AI Replace Accountants and Lawyers? 

Here, the conversation gets especially insightful. Nathan, who works with accounting and legal clients, asks an important question: if AI can already handle complex research and analysis, what role is left for professionals? 

Charles answers by pointing out something easy to miss: authenticity, accessibility, and trust aren’t things AI can copy. A firm that shows real experience, quick responses, and proven results still offers something a chatbot can’t match. But the standout point of the whole episode is this one, from Nathan: 

An accounting or legal firm carries professional indemnity insurance. An AI tool never will. 

If an AI gives bad advice, there’s no way to hold it accountable. But if a licensed advisor makes a mistake, there is accountability, and that peace of mind is valuable.  

This is exactly why account managers like Mary Aguirre Cruz and Katherine Durham, whose expertise has been built over years of hands-on client work, can’t simply be replaced by AI. That same relationship-first logic extends to how businesses show appreciation for their own clients.  

A thoughtfully chosen gift or giveaway is still one of the most effective ways to reinforce trust — it’s a tangible reminder that a client relationship is genuinely valued, not just transactional. That’s precisely why clients continue to order gifts even as more of their day-to-day work becomes automated. 

The “Hollowing Out” Theory 

The conversation’s central metaphor is the story of London’s black cab drivers. Before GPS, learning “The Knowledge,” or London’s street map, was a real challenge and conferred professional status on drivers. GPS didn’t just help with directions; it removed that barrier, making the job open to everyone.  

Applied to accounting and law, the theory suggests that routine research and lower-value bookkeeping work will increasingly be automated. Top firms, those who reinvest AI-driven efficiency gains into deeper advisory relationships, pull further ahead and can charge a premium. It’s a trend Nathan describes as concerning for the next generation entering these professions, since the traditional junior-to-senior career ladder may no longer look the same. 

Industries With a Built-In Moat 

Not every industry faces the same risks. The conversation shifts to construction, where Nathan says Australia has built-in protections that make disruption, whether from AI or other sources, much less likely: 

• Regulatory barriers — strict federal and state rules around approved building materials 
• Geographic isolation — Australia’s remoteness limits the kind of large-scale informal labour supply seen in other countries. 
• Strong union influence — labour supply stays tightly controlled. 
• Small, controlled labour pool — unlike markets with large undocumented workforces feeding the supply chain 

Charles points out that the downside of these protections is that they also keep housing costs high. This is a sobering thought for anyone hoping homes will become more affordable for the next generation. 

It’s also an industry where being physically present on-site is still very important. That’s why uniforms and hi-vis workwear are still a practical way for construction businesses to show their brand—people see them at work, not just online. 

What Endures 

Zooming out, the pair touch on which industries are likely to remain resilient regardless of how far AI advances: entertainment, food, and healthcare. People will always need to eat, be entertained, and be cared for — even as those needs are met, they continue to evolve. 

For businesses operating in these spaces, the opportunity is to lean into the physical, human side of the experience — think branded drinkware and picnic gear for food and hospitality brands, or comfortable apparel and wellness items for healthcare providers looking to stay memorable in a crowded market.