Generative AI tools have become common in professional knowledge work, but their value in business strategy is limited when used as disconnected conversations. The core bottleneck is not access to intelligence but the absence of structure connecting diagnosis, decisions, priorities, and execution.
Embedding AI into structured management workflows, rather than relying on isolated prompts, allows it to support continuity across strategy, planning, and review. This perspective informed the development of Cogliva, a tool designed to move leaders through connected strategic thinking rather than one-off AI responses.
Artificial intelligence has moved very quickly from being a future-oriented topic to becoming part of everyday knowledge work. Many professionals now use tools such as ChatGPT, Claude, Copilot, Gemini, and others to write, summarize, brainstorm, compare options, prepare documents, and think through complex questions. This is no longer limited to technology teams or AI specialists. It has become part of how managers, consultants, entrepreneurs, analysts, and experts work. This is a major shift.
But the more I have used these tools for serious business and strategy work, the clearer one thing has become to me: The bottleneck is no longer only access to intelligence. The bottleneck is structure. AI can generate answers. It can create drafts. It can challenge assumptions. It can explain complex topics. It can help us move faster. But business and strategy work are not only about isolated answers. They require continuity. They require context, diagnosis, choices, objectives, priorities, trade-offs, ownership, follow-up, and review. They require a way to move from a question to a decision, from a decision to a plan, and from a plan to ongoing management. A good AI conversation can be very useful. But leadership work cannot remain a collection of disconnected conversations.
The power and limits of general AI tools
I have written before about business strategy, digital transformation, artificial intelligence, prompt engineering, and the practical use of generative AI in business. I still believe these topics are deeply connected. General AI assistants are powerful because they lower the barrier to knowledge work. They help people start faster. They reduce the friction of writing and analysis. They can act as sparring partners, editors, researchers, summarizers, and idea generators. For many tasks, this is already transformative. But when we move into serious management and strategy work, the limitations also become visible.
A leadership team may ask:
- What is our real strategic challenge?
- Why is growth slowing?
- Which market opportunities should we prioritize?
- How should we translate our strategy into measurable objectives?
- Which initiatives should receive management attention?
- What external changes should influence our assumptions?
- How do we keep strategy alive after the planning workshop?
AI can help with each of these questions. But if every question starts from a blank chat, the work can easily become fragmented. The same business context has to be explained repeatedly. The assumptions are not always carried forward. The outputs may be good individually, but weakly connected. A diagnosis may not translate into strategy. A strategy may not translate into execution. A plan may not remain connected to external changes. This is not a weakness of AI itself. It is a design issue around how AI is used.
Prompting is useful, but it is not enough
Prompt engineering matters. A better question usually leads to a better answer. But a good prompt is not the same as a good management process. In business strategy, the quality of the work depends not only on what we ask AI to produce, but also on the path we follow before and after the answer. Before the answer, we need to frame the situation properly. After the answer, we need to make choices, connect implications, assign priorities, and revisit assumptions. This is where I think many organizations will face the next challenge.
The first phase of generative AI adoption was about discovering what AI can produce. The next phase will be about designing how AI fits into real work. For knowledge workers, consultants, executives, and entrepreneurs, the question is no longer only: “Can AI help me write this?”. The better question is: “Can AI help me think, decide, and manage better?”
From content generation to structured knowledge work
This distinction has become increasingly important in my own work. My professional background has for many years sat at the intersection of business strategy, digital transformation, applied AI, management systems, and organizational change. I have also written books and articles about online business, AI, and the changing nature of knowledge work.
Over time, I became less interested in AI as a tool for producing more content and more interested in AI as a way to improve the structure of business thinking. That shift matters. Because in many organizations, the problem is not a lack of documents. There are already many documents, presentations, reports, dashboards, templates, and meeting notes.
The harder problem is often that these pieces are not sufficiently connected. Strategy is written in one place. Risks are discussed somewhere else. KPIs are tracked in another tool. Market signals are noticed informally. Management actions are captured in meeting minutes. Business model assumptions live in people’s heads. Strategic decisions are revisited too late. AI can help, but only if it is embedded into workflows that reflect how management work actually happens.
The next phase: AI inside business workflows
I believe the next important step for AI in business is not only better models. Better models will come, and they will matter. But the bigger practical question is how those models are embedded into structured workflows. For strategy and management work, that means combining AI with:
- business context
- diagnostic logic
- management frameworks
- structured questions
- objectives and KPIs
- strategic assumptions
- external signals
- review cycles
- decision history
- execution follow-up
This is different from asking a chatbot to produce a strategy document. It is about creating a more continuous way of working, where AI supports the movement from understanding to decision, from decision to action, and from action to learning. That is also the direction that led me to build Cogliva.
Why I started building Cogliva
Cogliva started from a simple question: What would it look like if AI was not only used to answer business questions, but to structure strategy and management work itself?
I did not want to build another chatbot. I wanted to explore whether AI could support a more connected path through the work leaders and consultants already need to do. In Cogliva, the aim is not simply to ask AI for a business recommendation. The aim is to move through a structured flow: understand the challenge, diagnose the situation, design strategy, translate it into tactical plans, monitor relevant external changes, and keep the work alive in a strategy workbench.
This reflects a broader belief I have developed through using AI in real business contexts: AI becomes more valuable when it is not only conversational, but contextual, structured, and connected to action.
Of course, this is still a journey. Tools, workflows, and management habits will continue to evolve. But building Cogliva has helped me think more clearly about where AI may create real value in knowledge work.
Three lessons from building with AI for strategy work
There are three lessons that stand out for me so far.
1. AI is strongest when the business question is well-framed
Many weak AI outputs are not caused by weak AI. They are caused by unclear questions. In strategy work, the first challenge is often not to find the answer. It is to define the real problem. Is the issue growth, profitability, positioning, customer relevance, operating model, innovation, execution, or leadership alignment? Is the visible problem a symptom of something deeper? Is the organization trying to solve the right issue? AI can help explore these questions, but only if the workflow encourages proper framing before jumping to recommendations.
2. Frameworks still matter
AI does not remove the need for business logic. Strategy, business models, risks, KPIs, OKRs, management systems, market positioning, and execution still require structure. Frameworks are not perfect, and they should not be used mechanically. But they help create shared language, disciplined thinking, and comparability. The role of AI is not to replace frameworks. It is to make them more usable, more adaptive, and easier to apply in real situations.
3. The future of knowledge work is not only faster writing
Much of the early excitement around generative AI has been about productivity: faster emails, faster reports, faster summaries, faster presentations. That matters. But I think the deeper change is different. The deeper change is that AI can alter the operating model of knowledge work. It can change how we frame problems, how we prepare decisions, how we compare options, how we document reasoning, how we connect strategy to action, and how we revisit assumptions when the external environment changes. This is especially relevant for leaders, consultants, entrepreneurs, and experts whose work is not only to produce content, but to improve judgment.
What I will explore next
This is also the direction I want to explore more in this newsletter. In the past, I have written about business, technology, AI, digital transformation, and strategy. These topics remain important. But I now increasingly see them through one connecting question: How can AI help leaders and professionals do better knowledge work, not only faster knowledge work?
That question touches strategy.
It touches management systems.
It touches venture building.
It touches digital transformation.
It touches how organizations sense change, make decisions, and act.
Some of my reflections will come from my professional work. Some will come from my writing and research. Some will come from what I am learning while building Cogliva. But the central theme will remain the same. AI is changing knowledge work. The opportunity is not only to produce more. The opportunity is to think better, decide better, and manage better.