Hi , On-Balance Issue 176: AI and The Illusion of Excellence

Hello there and welcome back to On Balance, our fortnightly round-up of news, views and research around the issue of diversity, equity and inclusion.

AI and The Illusion of Excellence

When Micha Kaufman, CEO of Fiverr, warned about AI’s seismic impact on jobs last week, his advice on how workers should prepare sparked reactions from alarm to applause.

His message hit a nerve not just for its content but because it exposed something deeper: our collective uncertainty. AI is moving through organisations at speed, and we remain unsure and divided on how to respond.

That the warning came from Fiverr should make us pause. A company built on disrupting traditional work models now fears disruption itself. If giants like Google are grappling with how to maintain relevance, no one is immune to the curve of change.

Whether you see AI as a threat or a path to better work, its advance reveals something deeper: AI doesn’t weaken the value of human judgment—it strengthens it, if we know how to use it well.

AI’s appeal is obvious: speed, scale and simulated expertise. CEOs draft speeches without comms teams. Lawyers are replaced in writing court submissions. The outputs may look polished but can be 60% accurate and 40% dangerous. A speech unchallenged can miss the mark. Legal advice may draw on fiction - as seen recently in the US.

This isn’t a call to resist AI, but to recognise where human input is non-negotiable.

AI can research, draft, optimise and polish. But it doesn’t challenge assumptions. It won’t sense discomfort or warn when something might land badly. There’s no hand on your wrist or whisper in your ear asking: is this right?

For leaders, that creates a hidden risk: the illusion of competence without collaboration - and the even bigger illusion of self-awareness.

The real leadership challenge with AI isn’t the tech. It’s understanding how we, as humans, respond to it. That means looking inward. Leaders need to understand their own thinking patterns, cognitive biases and emotional triggers.

One of the unsettling things about working with AI is how quickly it mirrors your speech patterns - making it even more important to disrupt your own defaults and stress test its output. A simple tactic: ask for the counter argument.

I know I have a bias towards new - I get excited by innovation. But I have to counter that with self-questioning and turn to those who see the risks. AI can be a seductive thought partner, but it’s also sycophantic and subtly dangerous.

It’s tempting to surround yourself with people who just “get on with it.” But that is a false - risky - comfort. True leadership means embracing difference - of thinking, experience and style - even when it slows you down.

Leaders who don’t examine their reflexes - be it tech optimism, doubt or detachment - risk leading from assumption, not insight. The decisions may appear confident but lack the critical depth to hold up in the real world.

This matters more than ever because, like it or not, every team now has an invisible member: AI.

Unlike any other hire, it arrived with no onboarding, no role clarity and no job description. We’ve welcomed its output without redesigning how we work with it - or each other.

This isn’t a tech challenge. It’s a leadership one.

The hard part is not understanding AI but adapting the behaviours and expectations that shape its use. That includes asking: what assumptions am I making? What needs testing? Whose voice isn't in the room?

When we bypass the team in favour of the tool, we lose the mess - the questions, the friction, the ideas we hadn’t considered. It’s slower, harder and more human, but that’s where the value is.

Left unchecked, AI can fragment teams. Conversations become siloed. Some disappear into AI tunnels, others check out. Culture weakens.

Without clear boundaries, what looks like potential becomes an enfant terrible: technically brilliant, culturally destabilising, hard to rein in.

The leadership imperative now is to ground ourselves while shaping what’s next. To build new norms that keep AI in context and human judgment at the centre. To be clear on what’s possible - and honest about what’s real.

And crucially, to stay connected to purpose. The need to bring in diverse voices - of background, thought and lived experience - hasn’t gone away just because “DEI” has fallen out of fashion. We can’t afford to become deniers of that, either.

Good leadership in an AI-powered world still looks like good leadership in the one that came before it - just with more self-awareness, sharper thinking, and a renewed courage to listen well.