Smarter Tools, Same Questions: Why Human-Led Strategy Still Wins in an AI-Powered Business

For the last handful of years, AI has dominated business conversations. What can it do? Should we be leveraging it? What are the risks if we do? What are the risks if we don’t (the FOMO is real)? Boards continue to ask about AI. Leaders are finding new ways to experiment with it daily. Marketing departments are adopting generative & multimodal AI tools faster than anyone has time to evaluate them properly.

Somewhere in the middle of all of that noise, a quieter and more important question is getting lost:

What has to come first before any of these tools can actually move your business forward? And what still requires a human being with experientially grounded judgment, context, and consequential accountability?

At McCarthy Group, we think about this constantly. Not because we are skeptical of AI. We are not. we believe every business has at least one meaningful use case for AI whether they have identified it yet or not. We think about it because the businesses we talk to every day are navigating the same tension: the pressure to adopt AI quickly and the absence of a clear strategy for doing so in a way that is responsible, secure, and actually connects to their goals. Those are very different problems with very different solutions.

Some will argue that these are temporary problems; That as AI becomes more capable, it will absorb the strategic complexity too, learning your business deeply enough over time to surface the right questions, identify the misalignments, and recommend the right path forward. The most concrete version of that argument starts with memory and recall, which remain significant limitations on what AI can currently do with organizational context at any meaningful scale.

even if the AI “Memory and context Problem” is eventually solved, That's Not The End Of The Story

A few years ago, the conventional wisdom was that AI tools had little-to-no meaningful memory of your business. Every interaction with a conversational interface like ChatGPT or Claude started from scratch. An AI model had no record of the conversation about what your business tried last quarter, no awareness of decisions it helped you analyze six months ago, and no ability to carry context from one session to the next. For businesses hoping to use AI as a meaningful operational partner, that was a significant limitation.

The research and engineering communities have been working on this problem steadily, and the progress is real. In the early days of enterprise AI adoption, the dominant approach to giving AI systems access to organizational knowledge was a technique called retrieval-augmented generation, or RAG for short. The basic idea was to break large bodies of information into smaller pieces, convert those pieces into a mathematical representation, and retrieve the most relevant ones at the moment a question was asked. It was a reasonable workaround for the constraints of the time.

As models became more capable of holding larger amounts of information in their working memory at once, that workaround started to look less necessary for many use cases. By 2024 and 2025, the question of whether expanded context windows might eventually replace retrieval-based approaches had moved from academic discussion into active experimentation across the industry. The honest answer that seems to have emerged is that both approaches have genuine strengths depending on the situation, and that the more interesting question was not which method wins but how to give AI systems the right information at the right moment in a way that actually serves the business and is cost effective.

Today, experts like Andrej Karpathy are documenting practical approaches to organizational knowledge management (he refers to them as LLM Knowlege Bases) that look quite different from the systems built just a few years earlier. Instead of complex technical infrastructure, the emerging approach is closer to something a non-technical person might actually recognize: a well-organized, continuously updated library of information that an AI system can read, navigate, and reason across the same way a knowledgeable colleague might flip through a well-kept filing system. The sophistication is not in the retrieval machinery, rather, it’s in the curation. Someone still has to decide what goes into the AI knowledge base, how it is organized, and whether it stays current. The limitations that once prevented AI from holding meaningful business context over time are actively being addressed; but the strategic barriers, knowing what to capture, how to organize it, and what to do with it, remain entirely human problems.

So what does that mean for the argument that human strategic expertise still matters?

It actually strengthens it. Here’s why:

Whether the AI memory problem gets fully “solved” in two years or twenty, or whether it turns out to be more complicated than the current optimism suggests, the strategic argument holds either way because the memory & recall problem was never the only problem. It was just one of the more visible ones.

Even if AI systems eventually become capable of retaining and recalling organizational context with complete reliability, someone still has to decide what context matters. Someone still has to curate the knowledge that feeds the system, maintain it as the business changes, and ensure the strategy underneath all of it is actually sound. Someone has to ask whether the system is being built around the right goals in the first place. Those are not technical questions. They are strategic ones. Answers to business strategy questions don’t get easier as the tools get more powerful; they become more consequential.

Operational and growth bottlenecks do not disappear as AI memory improves and it becomes a more intuitive and valuable tool for businesses. The nature of the bottleneck changes.

The businesses that struggle will not be the ones that failed to adopt the right tools. They will be the ones that adopted tools without a clear strategy underneath them. That is where the human element becomes not just useful but necessary. For businesses navigating growth, operational complexity, and AI adoption simultaneously, that element rarely exists inside the organization in the form it is needed.

This is where an AI-fluent strategic consultant becomes one of the most valuable partners a growing business can have. A consultant who’s worth their salt does not just show up with a one-size-fits-all framework and disappear. They learn your business, learn your people, remember what was tried, understand what your experiences should mean for the decisions in front of you today, and ask the questions that are hard to ask from inside the organization. The goal is never dependency. It is to build the internal clarity and capability that allows your team to operate at a higher level long after the engagement ends. The relationship itself is the asset. And as the business grows and the questions get harder, there is a real person who knows the history, understands the context, and can help you think through what comes next.

That kind of partnership is what invaluable strategic work actually looks like.

Here’s the thing about advancing AI capabilities: they do not simplify the strategic problems businesses face. They expose them.

The Alignment Problem

Here is an example of a common strategy problem in today’s AI-powered world: companies often come to us not because they lack talent or tools, but because the talent and tools are not aligned. A marketing team executing beautifully toward the wrong goal. An operations team optimizing a process that should not exist. An AI adoption “strategy” that is really just a hodgepodge of subscriptions nobody has had time to properly evaluate against actual business objectives. A sales team with a strong pipeline and a conversion problem nobody has named yet. A product roadmap built around what the team can build rather than what the market actually needs. A customer success function that is reactive by design when the business model quietly depends on retention.

The value of outside strategic perspective is not that it replaces your team. It is that it sits above the day-to-day and asks the questions that are often difficult to ask from within the organization:

• Are the AI tools we are investing in mapped to a real business objective, or are we keeping pace with what everyone else seems to be doing and hoping it leads somewhere useful?

• Is our marketing strategy actually connected to our growth goals, or has it developed its own logic that nobody has stopped to interrogate? I.E. are we working in silos? Where?

• Is our sales process built around how our best customers actually buy, or around how we prefer to sell?

• Are we measuring the right things in operations, or just the things that are easy to measure? Do we have a go-to-market strategy, or do we have an array of tactics that have never been tied to a coherent market position?

• Is our customer success function set up to drive retention and growth, or to manage complaints?

• Does our leadership team have a shared definition of what winning looks like in the next twelve months, or does everyone have a slightly different answer? To that end, who “owns” our definition of winning?

Those questions require a strategic partner who is close enough to your business to understand it and far enough outside it to see it with clarity and objectivity. They require cross-functional experience that does not exist inside a single department or discipline. And they require the kind of candor that is genuinely hard to practice when your job, your relationships, or your standing inside the organization are all tied to the answer.

They are also the questions that become more important, not less, as AI systems become more capable.

A more powerful tool pointed in the wrong direction just gets you to the wrong place faster.

The Upskilling Problem

Misalignment between talent and tools is a frustrating problem. There is a related problem that tends to sit just beneath it.

The conversation we are having most often with leaders right now goes something like this: we have smart people, we bought the cutting-edge tools, we have told the team to use them. We are not seeing the results we expected.

The issue is almost never the tools and almost never the people. Nobody has taken the time to answer a more fundamental question first: what are we actually trying to accomplish, and which of these tools genuinely helps us get there? Without that clarity, even a talented team ends up experimenting in the dark. They adopt what is trending, use what is easiest, and measure what is convenient. None of which is the same as building real capability.

Helping a team use AI well is not a training exercise. It is a strategic one. It starts with understanding the business well enough to know where AI creates genuine leverage and where it is just adding complexity. That understanding rarely comes from inside the function doing the work. It requires someone with enough perspective to see the whole picture.

There is also a people dimension that tends to get skipped entirely. Adopting new tools and new ways of working is a change management challenge as much as a strategic one.

The businesses that get AI adoption right are rarely the ones that rolled out the fastest. They are the ones that brought their people along deliberately, addressed resistance early, established clear ownership, and treated adoption as an organizational shift rather than a software installation.

Strategy sets the direction. People make it real.

The Sustainability Problem

All of which raises a question worth addressing directly: if getting this right requires strategic clarity, cross-functional perspective, and careful attention to the people side of change, where does outside help fit in without becoming a crutch?

It is a fair question. The honest answer is:

a good engagement should make itself Less necessary over time.

At McCarthy Group, we say this out loud from the first conversation.

In practice, that starts with a deep dive strategy assessment that gets us both clear on where your business is, where it needs to go, and what is getting in the way. From there, most of the leaders we work with find the most value in staying in the work together through implementation, where the real strategic shifts actually happen.

For those who prefer to move independently, we make sure the plan we hand over after the strategy assessment is specific enough, grounded enough, and honest enough to actually be used.

The measure of a successful engagement is not how long it lasts. It is the distance your business covers toward its goals as a result.

What This Means Right Now

The through line of everything we have covered is this: AI is advancing faster than most businesses are building the strategic foundation to support it.

Most leaders we talk to are not short on ambition, talent, or resources. What they are looking for is someone who can help them cut through the noise, get clear on what actually matters, and build a path forward that accounts for where the business is today and where it is genuinely trying to go.

The companies pulling ahead are not necessarily the ones with the most sophisticated tools or the largest technology budgets. They are the ones who got clear on their strategy first, aligned their people and their tools around it, and built the internal capability to keep moving forward as the landscape shifts beneath them.

That is the work worth doing. Not because it is urgent in the way a quarterly target feels urgent, but because

the cost of strategic drift compounds quietly and becomes significantly harder to correct the longer it goes unaddressed.

Especially as AI systems become more capable of amplifying whatever direction a business is already heading.

Here is the question worth sitting with: is your business sufficiently clear on its strategy to sustain what it has built and capture what it wants to grow in a market that is evolving faster than many organizations can keep pace with?

Explore the answer to that question during a discovery call with McCarthy Group. No pitch, no pressure. Let’s have a candid conversation about where your business is and what it would look like to move forward with clarity and intention.

Schedule your free discovery call here.