Overcoming AI scepticism: The case for a Darwinian strategy

IN RECENT months, Singapore’s conference circuit – spanning innovation, leadership, governance, tripartite collaboration and digital transformation – has converged on a single question: What will artificial intelligence (AI) change in our professional lives and businesses?

Across panels and keynote sessions, I heard stories of bold new ventures, striking productivity gains and creative experiments that pushed artistic boundaries.

There were also the inevitable warnings – predictions of mass job losses and disappearing professions. Yet, during coffee breaks and side conversations, I noticed a different sentiment. Most of us were intrigued, but also a little bewildered.

The enthusiasm was tempered by a quiet scepticism about what AI could really achieve. And many wondered how to manage this in our own smaller or medium-sized organisations.

From curiosity to daily habit

Like many others, I’ve already woven generative AI (GenAI) into my routines. I’ve switched from Googling to ChatGPT-ing. In doing so, I’ve learnt that models like ChatGPT or DeepSeek are designed to always provide an answer – and rarely admit they don’t know. That makes them useful, but also risky.

Still, getting a coherent, natural-language response is often far more efficient than sifting through pages of sponsored links. Specialised AI tools help with my research and article preparation. These tools augment my work, making me faster and sharper – but they don’t represent the sweeping revolution often promised.

So, the real question becomes: How can small and medium-sized enterprises (SMEs) make AI truly transformative?

Automate, augment, add

AI offers three main opportunities. I call it the “Triple A” – it can automate jobs, augment employees, and also add entirely new business models.

Automation is the simplest to grasp. Text editing and translation, tasks once done by humans, are now handled seamlessly by AI.

Augmentation is more powerful. Take healthcare, for example. Radiologists review hundreds of scans daily, a process prone to fatigue and oversight.

AI-powered imaging tools can pre-screen these scans, highlight suspicious anomalies, prioritise urgent cases, compare historical data, and even generate preliminary reports. Yet, the ultimate medical judgment still rests with the doctor. AI doesn’t replace expertise; it amplifies and augments it.

And then there’s addition: AI-enabled business models that didn’t exist or weren’t possible before. Hospitality is shifting from standardised service to personalised experiences. Hotels once relied on fixed rooms and preset amenities. Personalisation was costly and reliant on human concierges who relied on memory or notes.

Today, AI analyses guest data – preferences, past stays, even social sentiment – to tailor stays. Virtual concierges offer 24/7 guidance, predictive systems adjust rooms before arrival, and dynamic pricing reduces vacancies. The result: hospitality that adapts to each guest.

Yet, many professionals still perceive AI as a foggy frontier filled with uncertainty, despite hearing of inspiring examples. Drawing from my research on project management under uncertainty, here are some recommendations to help SMEs operationalise AI.

Start with data discipline

The old adage still holds: “GIGO” – or garbage in, garbage out. An AI system is only as good as the data it is built on. Models will always produce an answer, even when the input is flawed. They will hallucinate if needed. To get valuable contributions from AI, reliable data is your first precondition for success.

That means analysing what trustworthy data you already have, ensuring privacy compliance, and structuring it for accessibility. Gather data from diverse, relevant sources and prioritise first-party data, that is, the kind your organisation owns.

Then, clean it relentlessly: remove duplicates, fix errors, standardise formats and normalise categories. Handle missing or outlier values carefully. For text or image data, apply proper preprocessing.

Embrace a Darwinian approach

Once your data is roughly 80 per cent reliable – perfection is a myth – move fast and experiment widely. Let the fittest ideas survive.

Importantly, don’t let your tech department monopolise AI initiatives. They play a vital support role, but the best applications will come from the front lines – from people close to customers or operations who spot real opportunities for improvement.

Such a Darwinian approach requires three actions.

First, build a culture of play and openness. Everyone in your organisation should understand what GenAI is and how it works. They don’t need to become data engineers or IT specialists, but short online courses can go a long way to create awareness. Encourage them not just to learn AI, but to play with it. Discovery should be exhilarating, not intimidating.

Second, bring the use of AI “above board”. Younger employees are likely already experimenting with it, often discreetly. Make that exploration visible and safe. As in education where we need to recognise that our students have constant access to AI applications, it is better to recognise and guide AI use than to ignore it. Encourage decentralised experimentation, but ensure transparency across the organisation.

Third, learn fast and scale fast. Not every experiment will succeed. Many will fail, and that’s fine. The key is to stop failures quickly and scale successes fast.

When a promising AI use case emerges, your IT and leadership teams must move swiftly to professionalise it: standardise processes, ensure security and integrate it into core systems.

This calls for a centralised learning capability – that is, a team responsible for evaluating, selecting and scaling AI applications. Speed of diffusion will be your competitive advantage.

From uncertainty to advantage

AI’s future may seem uncertain, but the path forward is clear: solid data, open experimentation, continuous learning as well as swift execution and diffusion. As the technology becomes ubiquitous, the true differentiator for SMEs is rigour – moving beyond curious exploration to embed the fittest ideas into the business to drive real value.

The writer is a professor emeritus of operations management and former president of Singapore Management University