Strategy #13: In the AI era, organizations must redesign their decision-making process.
There's one question I often ask in executive meetings.
"In your organization, are there people using this now?" Does AI help with any work?”
Almost every time, the answer is... "have" And it's often followed by a smile.
Then I asked further:
"And did decision-making within the organization improve as well?"
The room is quiet.
This is no coincidence; it's a widening gap in many organizations worldwide.
Organizations are using more AI, resulting in higher output. Some people seem less busy, while others seem busier (excluding those who never seem busy). However, decision-making remains slow, continues to produce the same errors, and the problems keep recurring.
The reason is that many organizations are undergoing Digital Transformation but have not yet started Decision Transformation.
And these two things are not the same thing.
What does Digital Transformation change, and what hasn't it changed yet?
Looking back over the past 10 years, most organizations have invested heavily in digital transformation.
We transformed paper-based documentation into a system, switched from in-person meetings to online sessions, turned data into dashboards, and introduced automation to reduce repetitive tasks.
These things are truly valuable and fundamentally essential.
But if you ask directly what Digital Transformation changes, the most direct answer is:
It changed. How the organization works.
But it didn't change anything. How organizations make decisions.
Many organizations have more data, more systems, and more tools, but their decision-making processes remain the same. They still use the same hierarchical structures, wait for traditional approvals, and there's a gap between the available information and better decisions that should be made.
already What changes is AI bringing?
AI isn't just adding new tools to organizations.
AI is asking questions that many organizations have never clearly answered. "Who made what decision, and based what decision on?"
If an organization doesn't yet have a clear answer to that question, introducing AI won't improve decision-making; it will only facilitate the decision-making process. It just gets faster.
And making decisions faster but not carefully enough may not be progress.
AI doesn't come in just one form, and that's something organizations need to understand.
Before discussing Decision Transformation, it's important to understand that the AI currently used by organizations falls into three vastly different levels.
Traditional AI This is a familiar type of automation, working well with tasks that have clear, repeatable rules and don't change much, such as data categorization, reading certain types of documents, or performing tasks according to pre-designed rules. This provides clear value, but it also has clearly defined limitations.
Generative AI AI, which many people rely on every day to help with writing, summarizing, translating, and thinking, has real benefits, but there are risks that I've found many organizations are unaware of.
As AI writes better, answers more confidently, and produces more credible output, people may start to believe that...
A good-looking output is a correct output…These two things are not the same.
Agentic AI This refers to a level where AI doesn't just respond or create content, but plans, performs multiple steps, coordinates processes, and in some cases, can execute certain tasks instead of humans, which is very exciting.
But as AI becomes more capable, the question of governance becomes an unavoidable risk.
The important question, therefore, is not what AI can do, but what type of work is suitable for which type of AI, and what decisions still require human judgment.
Hidden traps in the Thai language.
AI can translate quickly and accurately at a general semantic level, but the Thai language is not just about "meaning."
It has a tone, a level of politeness, a relationship between the speaker and the receiver, and implied meanings that change depending on the context of each situation.
AI may be able to translate sentences correctly word for word, but it may not understand what kind of sentence or tone of voice is appropriate to use in a given context—to which customer, which executive, or which partner.
The most common result is: The email read perfectly fine, but the recipient felt something was wrong because the tone didn't match the long-standing relationship they had.
The problem isn't with using AI to help with writing or translating.
The problem lies in using it without checking, reviewing, or incorporating the human context.
And if this happens with a typical email, imagine what happens when organizations use AI to help make decisions on matters with far greater impact.
What is Decision Transformation?
Decision transformation isn't about technology.
It's about organizations redesigning how data, systems, and AI will be used to make decisions, who owns those decisions, and how humans and AI divide roles in different types of tasks.
Simply put,
- Digital Transformation transforms an organization.work" how
- Decision Transformation changes how an organization...decide" how
Having a more powerful tool doesn't automatically mean better decision-making.
And from my conversations with executives in many organizations, what separates organizations that are truly successful with AI from those that just seem busier because of it isn't the tool itself, but the clarity on these four disciplines.
Discipline number one: Knowing which type of AI is best suited for which task.
Not every task requires the most sophisticated AI, and not every task should be automated by AI. Organizations that are clear on this will not waste time on tools that are not right for the problem.
Second discipline: Knowing how the output from AI needs to be checked.
Especially for jobs involving language, communication, customer service, law, finance, or high-impact decision-making.
The better the output looks, the more cautious you need to be, not less cautious.
Many people working in law, technical contracts, or finance understand this well. Documents written by AI may appear formal and complete at first glance, but certain words, conditions, or contexts can significantly alter their meaning.
The third discipline: Knowing who owns it and who is responsible for the decisions.
AI may help analyze, suggest alternatives, or warn of risks, but in many matters, humans still hold responsibility for the final decision, and that responsibility cannot be outsourced to AI.
The Fourth Discipline: Knowing when to use human judgment.
Because some things don't just require the fastest answer, but the answer that's most appropriate to the context, and these two things are often not the same answer.
Epilogue: The Era AI doesn't need humans who think less, but rather humans who think more clearly.
Many people fear AI because they believe humans will be replaced.
But what I've seen from organizations that are really using AI well is the opposite.
They aren't giving AI more decision-making power, but rather designing the system to enable humans to make better decisions, with AI as a tool.
The organizations that utilize AI best are not those that let AI do everything the most, but rather those that clearly understand when AI should be an assistant, when AI should be the analyst, and when humans must return to their original roles, irreplaceable.
The advantage in the AI era, therefore, is not in who has more tools, but in who has... Discipline in decision making better
From Digital Transformation, which brings organizations more systems and data, to Decision Transformation, which enables organizations to better utilize AI, data, and human judgment together.
This is the next step in transformation, and it's more important than any tool we choose to use.
And that leads to a question I'd like to invite you to think about together in the next episode…
If an organization wants “decision-making discipline” in the AI era… who within the organization needs to own that, and what kind of system will make it a reality?
