The people side of Artificial Intelligence

Log in or create a free account to get started.

Read History Go to My Account Log Out Enhanced by Google Opinion The people side of Artificial Intelligence Artificial intelligence is moving faster than most organisations can absorb it. New tools, agents, automation platforms and algorithms seem to

Listen to this article Share Artificial intelligence is moving faster than most organisations can absorb it. New tools, agents, automation platforms and algorithms seem to appear every week. Companies are buying licences, training employees in prompting, experimenting with agents and looking for ways to embed AI into everyday work. All of this is understandable. The productivity opportunity is real.

But I increasingly wonder whether organisations are investing as deliberately in the people behind the technology as they are in the technology itself. What happens when powerful AI meets poor judgement, weak culture and unclear accountability? What happens when employees know how to use AI but do not know how to think critically about what it produces? What happens when managers cannot decide when to trust an algorithm, when to question it and when to override it? And what happens when nobody is quite sure who owns the consequences of an AI-assisted decision? We saw all these in the social media revolution, and society is still paying dearly for this.

Perhaps the greatest obstacle to AI transformation is not AI. Management research has long reminded us that technology does not operate in isolation. Sociotechnical systems thinking argues that performance improves when technology and the human system around it are designed together. Changing the tool without changing roles, processes, relationships and decision rights often produces disappointing results. Herbert Simon and Daniel Kahneman similarly remind us that human judgement is bounded and vulnerable to bias, while Amy Edmondson’s work on psychological safety shows that people learn better when they can question, admit uncertainty and challenge assumptions without fear. These ideas become even more important in the age of AI.

The first challenge is capability. We are teaching people to use AI, but are we teaching them to think with AI? AI literacy is necessary. Employees should understand the tools available to them, how to prompt effectively and where AI can improve productivity. But AI literacy without thinking capability is not enough. AI can produce answers quickly; humans still have to determine whether those answers are useful, accurate, ethical and appropriate.

This requires what I like to think of as a human thinking engine built around critical thinking, creative problem-solving, and decision-making. Critical thinking challenges what exists. It questions assumptions, evaluates evidence, detects bias and verifies AI output. Creative problem-solving imagines what could exist. It helps people frame the right problem, generate alternatives and connect ideas in new ways. Decision-making then chooses what to pursue, weighing trade-offs, applying context and accepting responsibility for consequences. AI can augment every stage, but it cannot compensate for consistently poor thinking.

This is why I continue to use the Nigerian metaphor of the “Efiko” and the “Olodo”. AI is an amplifier. The Efiko uses it to learn faster, question assumptions, deepen expertise and solve harder problems. The Olodo uses the same technology to avoid learning, avoid thinking, copy and paste and produce mediocrity faster. The future will not belong merely to people who use AI. Almost everyone eventually will. It will belong to those who know how to think, learn and create with it.

The second challenge is judgement. Better prediction does not automatically produce better decisions. The objective should not be to choose between human judgement and artificial intelligence. It should be to design better human-plus-AI decision-making. That means organisations need clearer rules about when people should follow, question or override AI. Follow the system when the model is validated, reliable and operating within clear boundaries. Question it when assumptions, data quality, context or consequences are uncertain. Override it when legitimate professional judgement identifies something material that the system has missed. But an override should not mean simply ignoring the machine. It should create a feedback loop: reason, document, review and learn. If professional judgement departs from AI, the reason should become visible and learnable. Otherwise, organisations will never know whether the human corrected the machine or merely overruled it out of habit, ego or convenience.

The third challenge is organisational readiness and culture. The same AI can produce very different outcomes in different organisational environments. Culture, trust and accountability determine much of the difference. Can people question outputs, experiment and admit uncertainty? Do employees and customers believe AI is being used responsibly? Who owns the decision and its consequences? These are not technical questions, but they determine whether the technology creates value. Learning cultures are more likely to use AI as a thinking partner. Fear-driven cultures may resist it, ignore it, use it superficially or treat it the same way they do their bosses – as an oracle that can never be wrong. Accountability matters because AI risk increases when responsibility is fragmented across IT, vendors, business units, compliance and management.

This suggests a simple AI readiness test built around three conditions: capability, culture and clarity. Can our people work intelligently with AI? Can they question, learn and experiment? Do they understand the boundaries, decision rights and accountability surrounding its use? AI adoption is a technology decision. AI readiness is an organisational capability. That distinction matters because many organisations are still approaching AI as though successful transformation depends mainly on acquiring better tools. Tools matter, but sustainable value comes from something broader: people who can think better, decision systems that help them decide better and organisational environments designed to help everyone organise better.

So invest in the technology. Buy the tools. Train people in AI literacy. Automate processes. Experiment with agents. But build the people too. Because the more powerful artificial intelligence becomes, the more important human intelligence becomes.

Omagbitse Barrow is the founder of Efiko Management Consulting, and he supports organisations and leaders to translate their strategy to results.