Reading the recent coverage of AI and autonomous agents from the AFR Business Summit got me thinking. There’s wisdom in much of what was discussed – but something fundamental seems missing from the conversation.

The idea that AI will transform the “messy, mediocre middle” of our industry makes sense conceptually. Those templated emails, standardised designs, and formulaic customer journeys will certainly evolve. And that evolution brings opportunities.

But there’s a significant difference between acknowledging AI’s transformative potential and accepting the notion that autonomous agents will completely reshape businesses within a convenient six-month window.

Let’s consider this more carefully.

The emperor's new algorithms

I’ve spent 20 years in CX trenches. I’ve witnessed the dot-com bubble, social media revolution, mobile transformation, and now I’m watching the AI gold rush unfold with a sense of perspective that comes from seeing several technology waves.

Here’s what stands out to me: the biggest innovations in customer experience rarely come from technology alone. They come from deeply understanding human behaviour and designing around it.

Remember when everyone needed an app? Then a chatbot? Now it’s an autonomous agent. The technology evolves, but the pattern remains consistent.

Let’s be candid about what’s happening here. The big consultancies and tech vendors are racing to position themselves as essential navigators of this new AI landscape. When global firms talk about AI transforming the industry, they’re working to establish themselves as the gatekeepers before clients realise the full potential of these technologies might be accessible through alternative routes.

The interesting paradox? Many of these firms are themselves facing the very disruption they’re helping clients navigate.

The gap between AI capabilities and implementation

Let’s consider some technical realities. Discussions of “autonomous agents” raise implementation challenges.

These systems still suffer from hallucinations and factual inconsistencies – imagine an agent misinforming superannuation customers about withdrawal regulations or tax implications. The liability should make any financial services CMO sweat.

LLMs also struggle with extended conversations across multiple touchpoints. That effortless handover from chat to booking? It often relies on extensive guardrails, human oversight, and carefully constrained domains.

Integration is another hurdle. One tier-1 retail client spent 80 per cent of their AI budget not on fancy algorithms, but on backend plumbing to make their agents useful.

The technology is impressive, but the gap between polished demos and the messy reality of enterprise deployment is vast. Anyone claiming otherwise is selling something.

From Lucky Country to Lazy Country?

Most Australian businesses implementing AI aren’t doing anything remotely innovative – they’re following Silicon Valley templates created by people with limited understanding of our culture. By implementing these systems without questioning their underlying assumptions, we risk dependency – importing not just technology but also American values, decision-making frameworks, and perspectives embedded in code we rarely scrutinise.

The leaders who will thrive aren’t the ones who can buy the shiniest AI – they’re the ones who can clearly articulate which parts of their customer experience must remain fundamentally human, and which can be thoughtfully augmented by technology.

The human difference

Last week, I watched one of our designers spend hours understanding the emotional journey of someone navigating their superannuation fund after a major life change. The empathy, intuition and emotional intelligence she brought to that process can’t be replicated by an AI, no matter how sophisticated.

Yes, the forms she designed will likely be processed by automation. The confirmation emails might be drafted by AI. But the moments that matter – the points where customers are making life-altering financial decisions, feeling vulnerable, or simply overwhelmed by choices – those demand human touch. After all, human capabilities are the foundation of meaningful customer relationships.

A middle path for the future

So where does this leave the “middling middle” industry leaders claim is doomed? I’d suggest a more nuanced view.

The middle isn’t going away – it’s transforming. The formulaic parts will indeed be automated. But there’s a new middle emerging: people who excel at the human-AI intersection. Those who can prompt effectively, curate intelligently, and know when to override algorithmic suggestions with human judgement.

I recently interviewed a strategist who described herself as an “AI whisperer”. She wasn’t a programmer or data scientist. She was someone who understood both customer psychology and how to guide AI tools toward genuinely useful outputs. That’s the new middle, and it’s anything but mediocre.

The hard truths

For CMOs and CTOs navigating these waters, I’d offer these observations:

1. The best AI implementation is invisible to customers. When your AI becomes the focus rather than the customer experience, it may be worth reconsidering the approach.
2. AI amplifies what’s already present. Clear data streamlined processes and focused strategy become more powerful with AI. Challenges in these areas become more apparent too.
3. Technical debt affects AI implementation. Legacy systems that work well on their own may present new challenges when integrated with AI initiatives.
4. Culture shapes technology success. The most sophisticated algorithm exists within your organisation’s culture and service philosophy.
5. Consider how vendor interests align with yours. Each proprietary system you adopt creates a relationship that deserves thoughtful evaluation for long-term partnership potential.
6. Board-level understanding matters. Technical investments benefit from leadership that can thoughtfully evaluate both the opportunities and limitations.

Importantly, six-month timelines don’t reflect implementation reality. Thoughtful implementation takes longer – and that’s okay. Excellence matters more than speed.

What keeps me up at night

In an industry experiencing significant transformation, what keeps me engaged isn’t concern about technology replacing human roles. It’s ensuring that in our enthusiasm for AI, we remember what makes our work meaningful: helping businesses create genuine human connections with their customers.

When implemented thoughtfully, AI enhances our human capabilities. It can manage the repeatable so we can focus on the distinctive. It can process the predictable so we can address the personal.

Technology without human insight has limited potential. And human creativity without technology misses opportunities in today’s landscape.

Australia can be more than the Lucky or Lazy Country. We can be the Thoughtful Country, implementing technology wisely rather than rushing to adopt it.