← All insights

AI Won’t Save You — But It Will Change How You Work

Companion to The Uncommon Advisor, S1 E2

Dr Stephen Anthony, Avneet Singh Riat, Meehir Patel

Artificial intelligence has arrived in a way few technologies ever have. Not gradually, not quietly, but visibly—suddenly accessible, widely discussed, and immediately useful. It has moved from being a specialist capability to a mainstream tool in an incredibly short period of time. And with that visibility has come both excitement and anxiety in equal measure. Across industries, organisations are asking the same questions: How do we use it? Where does it add value? And perhaps most importantly—what does it mean for the way we work?

These are not abstract questions. They are being explored in real time, often without clear answers. As Meehir Patel reflects, even those working closest to the technology are “peeling back the layers” as they go, learning through application rather than certainty. That is the reality of AI today. It is powerful. It is evolving. And it is not yet fully understood.

AI Is Not New—But Its Accessibility Is

It is easy to think of AI as something entirely new, but as Avneet Singh Riat points out, technology-enabled change has been part of consulting and business transformation for decades. Automation, outsourcing, data processing—these are not new ideas. What has changed is accessibility.

AI has moved from specialist systems into the hands of individuals. It is no longer confined to large organisations or technical experts. It is now a tool that anyone—from a startup founder to a corporate executive—can access instantly.

This shift fundamentally changes the conversation.

It creates a world where:

• Analysis can be accelerated
• Content can be generated rapidly
• Processes can be streamlined almost immediately

And yet, it also creates a new tension. Because while capability has expanded, understanding has not always kept pace. As Avneet notes, we are still in the early stages of truly understanding what AI can do—and, just as importantly, what it should do.

The Reality of Using AI in Practice

One of the most important insights from the conversation is that the use of AI is not uniform. It is highly situational, shaped by context, risk, and purpose. In some areas, its impact is immediate and tangible. Meehir describes how AI has transformed parts of the proposal and bid-writing process. Tasks that once required coordination across multiple people and extended timeframes can now be accelerated significantly. AI can synthesise information, structure content, and provide a starting point that speeds up delivery.

Avneet reinforces this from a commercial and operational perspective. The process of building proposals—historically dependent on multiple inputs from different specialists—can now be streamlined, reducing effort while increasing speed. But what is striking is what they don’t do. They don’t hand over control entirely. Instead, they guide the process. They refine outputs. They validate results. The tool accelerates—but it does not decide. This distinction is critical. AI enhances capability. It does not replace judgement.

Where AI Should Not Be Used—At Least Not Yet

If one side of the story is about acceleration, the other is about restraint. Avneet’s experience in board-level engagements provides a clear example. In highly sensitive, confidential environments—particularly where strategy, growth, and commercial positioning are being shaped—AI was deliberately not used. The reasons are instructive. These environments require:

• Trust
• Nuance
• Interpretation of human dynamics
• Real-time interaction and judgement

AI, in its current form, is not equipped to operate effectively in those spaces. It cannot sit across from a leadership team and navigate the complexity of a conversation. It cannot interpret tone, political dynamics, or emotional undercurrents. It cannot respond to subtle shifts in direction. As Avneet puts it, it is not yet capable of handling multi-dimensional, complex problems that involve both people and strategy.

This highlights a broader point: just because AI can be used does not mean it should be used everywhere.

The Illusion of Certainty

One of the more subtle risks of AI lies in the confidence it creates. Outputs are often well-structured, coherent, and persuasive. They appear authoritative. But as Stephen Anthony cautions, that authority can be misleading. AI generates responses based on patterns in data. It does not “know” in the way humans do. It does not validate truth—it reflects probability. Stephen gives a simple but powerful example: if enough people assert something incorrect, AI may reproduce that assertion as if it were valid.

This becomes particularly important in analytical or strategic contexts. If organisations begin to rely on AI outputs without scrutiny, they risk making decisions based on incomplete or flawed reasoning. That is why, as both Stephen and Avneet emphasise, there is still a need for rigorous validation. At present, AI outputs are often subject to what Stephen describes as “100% inspection.” In other words, the speed has increased—but the responsibility to verify has not diminished.

The Rise of a New Skill: Prompting, Structuring, and Guiding

Another emerging theme is that using AI effectively is, in itself, a skill. Stephen highlights this clearly when he talks about the importance of writing prompts. It is not enough to simply ask a question. The quality of the output depends heavily on how the input is structured. This introduces a new dimension to professional capability.

It requires:

• Clarity of thought
• Understanding of audience
• Ability to structure information effectively

Meehir expands on this by describing the shift from simple prompting to more advanced approaches, including agent-based workflows and iterative refinement. But even here, a pattern emerges. The individuals who get the best results from AI are not those who rely on it blindly—but those who already understand the problem they are trying to solve. AI amplifies thinking. It does not replace it.

Speed vs Trust: The Tension at the Heart of AI Adoption

One of the most interesting dynamics explored in the conversation is the tension between speed and trust. AI clearly accelerates processes. It reduces effort. It enables rapid iteration.

But at the same time, it introduces uncertainty. Stephen captures this in a simple observation: organisations are using AI to produce outputs faster—but still checking everything manually because they do not fully trust it yet. This creates an interesting paradox. The tool saves time—but only up to a point. Beyond that, human validation re-enters the process, limiting the efficiency gain. Over time, as the technology improves, that balance may shift. Trust may increase, and the need for oversight may decrease. But for now, the human layer remains essential.

Different Organisations, Different Approaches

AI adoption is not happening uniformly. It varies significantly depending on the type of organisation. Large corporations tend to approach AI with caution. Governance, risk management, and control processes shape their adoption. There are concerns around data security, compliance, and reputational risk. At the same time, they are acutely aware of the competitive pressure. As Stephen notes, many organisations feel that if they do not embrace AI, others will—and that creates urgency.

Smaller organisations and startups, by contrast, often move faster. They see AI as a way to reduce cost, increase efficiency, and scale operations with fewer resources. This creates a divergence. Some organisations optimise for control. Others optimise for speed. The role of the advisor, increasingly, is to help navigate that balance—ensuring that adoption is both effective and responsible.

AI in Training, Learning, and Capability Development

One of the more interesting areas of impact lies in training and assessment. Stephen raises a fundamental challenge: if AI can generate coursework, complete assessments, and produce structured responses, how do we measure real capability? This is not just an academic question—it goes to the heart of professional development.

The traditional model of:

• Exams
• Written assignments
• Portfolio-based assessment

becomes increasingly difficult to rely on. In its place, there may be a shift toward more discussion-based, experiential assessment—what Stephen describes as “professional discussion,” where individuals are evaluated through conversation and demonstrated understanding This reinforces a broader trend. As AI takes on more of the structured, repeatable tasks, human value shifts toward interpretation, communication, and application. In other words, capability becomes less about producing outputs—and more about understanding them.

Why AI Does Not Replace Advisors

This brings us back to the central question. If AI can generate insights, analyse data, and accelerate workflows, why do organisations still need advisors? Avneet answers this directly by pointing to the complexity of real-world engagements. Strategy, growth, organisational design—these are not linear problems. They involve people, systems, trade-offs, and uncertainty. AI can support parts of that process, but it cannot own it. Stephen adds another dimension—one that is often overlooked. Leadership is, at times, a lonely position. Decision-making carries weight. Uncertainty creates pressure. In those moments, what leaders often need is not just information, but conversation. They need someone to challenge them, to support them, and to help them think clearly. AI cannot fulfil that role. It cannot build trust. It cannot create psychological safety. It cannot act as a “critical friend” in the way a human advisor can.

A Tool, Not a Substitute

If there is one idea that runs consistently through the discussion, it is this: AI is a tool. A powerful one, certainly—but still a tool. As Meehir reflects, it does not have experience, judgement, or the ability to navigate relationships. It can produce outputs, but those outputs only have value if someone understands and validates them. This framing is important. It positions AI not as a replacement for human capability, but as something that enhances it. The organisations that will succeed are not those that replace humans with AI—but those that combine the two effectively.

Final Reflection: The Next Chapter Is Still Being Written

If there is one conclusion to draw, it is that we are still early in this journey. AI will continue to evolve. Its capabilities will expand. Its applications will deepen. But the fundamental challenges organisations face—complexity, uncertainty, human dynamics—will not disappear. As Avneet suggests, this is a conversation that will need to be revisited again and again as the technology matures. For now, the role of the advisor remains clear. Not as the owner of answers, but as the guide who helps organisations navigate them. AI may change how we work. It may change how we think. But it does not remove the need for judgement, experience, and human connection. And for that reason, it won’t save you.

But used well, alongside the right thinking and the right guidance, it just might help you move faster, think better, and make more informed decisions than ever before.