#489 AI, Clear Thinking and Better Decisions: What Business Leaders Can Learn from Philip Topham
AI, Clear Thinking and Better Decisions: What Business Leaders Can Learn from Philip Topham
Artificial intelligence is rapidly becoming one of the most powerful tools available to business leaders. Yet many organisations still struggle with a fundamental question: how should leaders actually think with AI rather than simply use it as a tool for quick answers?
In a recent episode of the Leadership Podcast and Leadership Videocast, leadership expert Niels Brabandt spoke with Philip Topham, author of the book "CRAFT Thinking: A Playbook for Clear Thinking and Better Decisions with AI". The conversation explored how executives and decision-makers can move beyond superficial prompt usage and instead use artificial intelligence as a structured thinking partner.
The discussion between Niels Brabandt and Philip Topham highlights a critical shift in how leaders should approach AI. The key insight is simple yet powerful: artificial intelligence should not merely accelerate decisions. It should improve the quality of thinking that leads to those decisions.
Why Speed Is Not the Same as Better Thinking
For decades, digital transformation in business has largely been driven by speed. Faster supply chains, faster communication, faster data processing and faster decision cycles became the hallmark of modern management.
However, as Philip Topham explains in the conversation with Niels Brabandt, artificial intelligence introduces a fundamentally different dynamic. AI is not merely a tool that executes commands faster. It is interactive. It responds, challenges assumptions and can be used as a structured thinking partner.
This interactive capability changes the nature of decision-making. Rather than rushing towards the fastest possible answer, leaders can use AI to explore alternatives, challenge their own reasoning and identify blind spots before major decisions are made.
In other words, better decisions often require slower thinking before faster execution.
The CRAFT Thinking Framework
One of the central contributions of Philip Topham’s work is the CRAFT Thinking framework, which offers a structured method for interacting with AI systems.
Instead of treating AI tools like traditional search engines, leaders should provide structured context and guidance. The CRAFT framework encourages users to clarify several elements before expecting meaningful output from AI systems.
Context is essential. AI tools require clear information about the organisation, the situation and the strategic objective.
Role defines how the AI should respond. Just as a new employee requires clear responsibilities, AI performs better when its role in the task is defined.
Actions specify what type of work the AI should perform, whether analysing information, generating strategic alternatives or reviewing content.
Format determines how the result should be presented, ensuring the output is usable in a real business environment.
Target focuses on the intended audience or outcome, which significantly influences the quality of the response.
This structure transforms AI from a simple tool into a collaborative thinking assistant.
Why Many Organisations Misuse AI
A recurring theme in the conversation between Niels Brabandt and Philip Topham is the misconception that AI should deliver perfect answers instantly.
Many professionals approach AI systems as if they were advanced search engines. They ask a single question and expect a fully developed answer immediately.
However, effective AI interaction requires dialogue. It requires refinement, clarification and iterative thinking.
Philip Topham uses a powerful analogy. AI should be treated like a new intern who has arrived on the first day of work with no understanding of the organisation. Without context, guidance and clarification, the intern cannot deliver meaningful work.
The same principle applies to artificial intelligence.
Understanding AI Errors and Bias
Another important topic addressed by Niels Brabandt and Philip Topham is the issue of AI errors and so-called hallucinations.
Artificial intelligence systems operate on probabilistic language models. They generate responses based on patterns in data rather than verified factual knowledge.
As a result, inaccuracies can occur. However, these inaccuracies often reveal something valuable about the underlying language patterns or assumptions in the data.
For business leaders, this means AI should not replace critical thinking. Instead, it should become a tool for testing assumptions, identifying potential biases and exploring alternative perspectives.
Pressure Testing Decisions with AI
One of the most practical insights from Philip Topham’s work is the concept of pressure testing business decisions using artificial intelligence.
He describes three stages of thinking that leaders can apply.
Hindsight focuses on retrieving information and factual knowledge.
Insight involves challenging assumptions and testing the robustness of an idea. Leaders can ask AI to simulate difficult stakeholders, raise objections or identify hidden risks.
Foresight explores future scenarios and second-order consequences. Leaders can use AI to imagine how a strategy might evolve in a rapidly changing environment.
This structured approach allows organisations to use AI not just for efficiency but for strategic thinking.
How Organisations Should Begin Using AI
For organisations that have not yet integrated AI into their daily workflows, Philip Topham offers simple but powerful advice.
The first step is to experiment. Leaders and employees must give themselves permission to explore the technology without expecting immediate perfection.
The second step is to understand how AI relates to the organisation’s specific business model, industry and competitive environment.
The third step is gradual integration. Once leaders understand the capabilities of AI, they can begin applying it to marketing, strategy development, presentations, research and operational workflows.
Artificial intelligence becomes most valuable when it is embedded into everyday thinking rather than treated as a specialised tool.
Leadership in the Age of AI
The conversation between Niels Brabandt and Philip Topham ultimately reinforces a crucial leadership principle.
Technology alone does not improve decision-making. Better thinking improves decision-making.
Artificial intelligence offers leaders a new form of cognitive collaboration. Used correctly, it can challenge assumptions, test ideas and sharpen strategic thinking.
But the responsibility for clarity, structure and judgement remains with the leader.
For decision-makers in business, the message is clear. Artificial intelligence will not replace leadership. It will raise the standard of thinking required from leaders.
And those who learn to think with AI rather than merely using it will gain a decisive advantage.
Niels Brabandt
---
More on this topic in this week's videocast and podcast with Niels Brabandt: Videocast / Apple Podcasts / Spotify
For the videocast’s and podcast’s transcript, read below this article.
Is excellent leadership important to you?
Let's have a chat: NB@NB-Networks.com
Contact: Niels Brabandt on LinkedIn
Website: www.NB-Networks.biz
Niels Brabandt is an expert in sustainable leadership with more than 20 years of experience in practice and science.
Niels Brabandt: Professional Training, Speaking, Coaching, Consulting, Mentoring, Project & Interim Management. Event host, MC, Moderator.
Podcast and Videocast Transcript
Niels Brabandt
Making decisions with AI. And you probably now think, "Well, I think I can do a prompt or two, and then probably I have the decision." Well, is that all? We have an expert with us here today. Hello and welcome, Philip Topham.
Philip Topham
Hello, welcome. I'm happy to be here and from across the pond, as it were.
Niels Brabandt
Yeah, greetings to California. You've just written an amazing book which deals exactly with what we're going to talk about today: Craft Thinking, a Playbook for Clear Thinking and Better Decision with AI.
Niels Brabandt
So first, of course, I have to ask, because we all know that probably the speed of light is not quite as fast as how quickly people think that they are experts on AI. I, of course, background-checked your whole CV, so you are working in the industry on a very high level for a very long time now. However, it's still quite a lot of work to get these books out there. What was your motivation to do this book?
Philip Topham
Great question. So I had saw ChatGPT hit the scenes, and I kind of dismissed it. It sort of was a toy. And I thought it was pretty neat, worked in machine learning and things. And a few months later, it was 100 million downloads. And I said, "Huh." I think there's a "huh." There's a "there" there.
Philip Topham
And it got me reading and looking on YouTube, and I'm listening to all these experts, and they're telling me, "This algorithm, this prompt," and I'm going, "This doesn't make sense to me." So I committed myself to writing two articles a week for 50 weeks, so 100 articles. I just wrote my 195th article on Substack, and it was to clarify my understanding of AI, not the impact.
Philip Topham
And that's what the book became. I wrote it. A few millennia, I put some articles out. Some millennials said, "This is simple. I understand what you're doing. It's helping me understand AI. It needs to be a book." Then I wrote the book.
Niels Brabandt
Brilliant. Brilliant. And of course, I read your book now, and I saw that one aspect which would probably trigger some people where you say, "Why smarter isn't faster." When you talk to boards today, most people say, "Look, this is all about doing things faster, isn't it?" What do you mean with, "Why smarter isn't necessarily faster?"
Philip Topham
So for the last 50 years since computers have been around, it's been speed, speed, speed, fast, fast, fast, go quick, efficiency, make that supply chain, ship overnight. Those are wonderful things for efficiency.
Philip Topham
But this is the first tool that talks back to you. And I'm talking about the interactive ability to ask chats questions and challenge your own thinking, work. That's what when you're building something fast, you don't want to run down a blind alley. You don't want to run off a cliff. You want to be deliverable about where you're going and what you're doing.
Niels Brabandt
Excellent. And you probably know some people who said, "Oh, I tried to use AI, and look, the results when I checked them, they were basically not too precise." And you have excellent chapters on stepping into the right shoes and go for specifics. How can you just put this into words where people say, "This is relatable to me?" Because many people say, "I just want to sit there, chat into it, and then get a brilliant result on attempt number one."
Philip Topham
Yeah, perfect. So a lot of people do use it like they've been used to using search engines, and they type in a question, and they get an answer, and they go, "Oh, but it wasn't the answer I wanted." They try again. But with chat, for some reason, they ask the question, they get an answer, and it's like, "Oh, hum."
Philip Topham
I like to think of chat as an intern that shows up to day one of your job. They've applied for the job, but they have amnesia. They don't know anything about the company. They don't know what job they're doing. They don't know anything. So if you approach it from that perspective, it doesn't know anything.
Philip Topham
You have to provide all the conversation, all that background of who you are, what you're doing, marketing, whether we're talking about marketing or aerospace engineering or drug discovery, that's where the questions start. Now it can orient itself and answer much, much better questions. And we can dig deeper into that as well.
Niels Brabandt
Excellent. And you created a framework, the Craft framework. Can you tell us a bit more about that?
Philip Topham
Yeah. So Craft was so I don't like the term prompt engineering because it's really about structuring the way you're thinking about things. And so if you're going to talk to this intern, you're going to tell them what's going on. You're going to give them the context. This is my company. This is what we do here. This is how big we are.
Philip Topham
"Oh, by the way, your job, your role, your role is you're going to be in the marketing department. And these are the types of actions or tasks that we're going to ask you to do. You're going to do marketing copy and review and make sure it's on brand tone. And then this is the format. We're working on Instagram. We're working on YouTube. And
Philip Topham
the target, what's the final T is the target. Who's it for? What audience are we? It's very different if you're targeting millennials versus baby boomers or some other generation. So it's a very simple framework, but it's more to organize our own thinking than it is about prompting.
Niels Brabandt
Excellent. And when we now try to go beyond the first answer, I give you an example which just happened to me. I watched the movie Marty Supreme, and I wondered, is this autobiographical? Is this about a real person? Is it biographical from someone? So I ask a certain AI tool - I'm not going to name it here - is the movie Marty Supreme biographic? It's about a table tennis player. Did this person really exist?
Niels Brabandt
And it says, "Yes, it is. It deals with a person that was in the circuit for quite a while whose name is John McEnroe." So fortunately, I am of the generation so maybe we are quite close from the I know that John McEnroe did a lot of things, but he definitely did not play table tennis. So I said, "Look, I don't think that this can be true because John McEnroe played tennis but not table tennis. Could you please recheck?" And then it said, "Oh, yeah, correct. It's actually about that kind of American table tennis player, and that person actually really existed."
Niels Brabandt
So the question is, many people say, "If I have to background-check everything by myself, I can actually do the whole work in the first place." And also, many people are very insecure about how to spot blind spots and biases. Can you give any advice on that one?
Philip Topham
Absolutely. So that's very interesting. So I would never call ping-pong table tennis. So this is something very common, and I use this as context. And it's something we naturally do as humans, which is called ambient awareness or situational awareness.
Philip Topham
I live in Southern California, and if I told people, "I'm going to go to the river this weekend," they would automatically understand, "I'm likely going to go to Lake Havasu and jet ski or motorboat around that place." If I was living near New Orleans and said I was going to go to the river, they would go, "Oh, you're going to the Mississippi River." They'd automatically make that connection.
Philip Topham
When you use table tennis, the AI is thinking more along the lines of tennis. You've got to remember, these corpuses or these language models are probabilistic. So the tennis probably means tennis, not ping-pong, right? Table tennis, when you say, "Oh, it's ping-pong," that's more likely how it was described in the papers.
Philip Topham
So this is why this relationship is really important, and particularly for countries - this is a big topic - countries have their own culture, history. They need their own language models to preserve their country's history and legacy. Otherwise, it gets a mosh pit muddled. Does that help clarify that?
Philip Topham
It's perfectly normal. And this is where, if you understand how it works, it's not a frustration. It's like a gift because it's telling you what the average language feels like. And it's a gift for marketing to understand, "I need to be more precise in my language."
Niels Brabandt
Excellent. And when it now comes to using that on a larger scale, maybe you say, "I'm a mid-sized business," or, "I'm a larger business, want to implement AI." How can I do something which you in your book call pressure testing with AI?
Philip Topham
Yeah, that's fantastic. Yeah. So there's three phases to sort of really thinking. One is called
Philip Topham
hindsight, insight, and foresight. So hindsight is more like you're looking for knowledge, like, "Who's the best table tennis player?" It's a very discrete fact.
Philip Topham
And then insight is you're trying to pressure test something. You've got an idea, "Hey, I'm going to go into an important meeting. There's going to be a salesperson there and the CFO. This is what I'm going to say in my presentation. Help me role-play, pressure test my assumptions. This is my plan. Be the worst possible CFO. Ask the most difficult questions of me." And they'll do that very nicely.
Philip Topham
And you can pressure test your own biases, and you can say, "What's my hidden flaws? What's my hidden biases? What I didn't think about? What's the oddball question I'm going to get?" That's pressure testing, insight.
Philip Topham
And then what a lot of companies don't think about is forward-thinking. We're in this hyper-change. And so I often ask companies to say, "You've got a new company. You've got a company. You're worried about AI. Well, what do you think about it? What would an AI-native company look like?"
Philip Topham
And then you come up with assumptions, and then you go, "Well, gee, what does that look like if it's 10 or 100 times bigger or faster or more capable? What does the future look like? What's the second, third, fourth-order implications?" So that's how I walk through the different modes of thinking, pressure testing, tell me where I'm wrong.
Niels Brabandt
Brilliant. And when now people say, "Okay, look, I think this is all perfect. However, I really don't know where to start," when people say, "I think I have a small business. I think I want to roll it out to my people. I just don't know if I can just tell them, 'Here it is,' and then use it, just try to make the best out of it," what would you say are the first one, two, three steps if someone says, "I want to implement that in my business, in my world?"
Philip Topham
Yeah. First step is give yourself permission to play and give yourself permission to type something in and get pissed that it gave you a bad answer, right? Because it's going to give you a table tennis answer, right? But it's not the computer's fault. It was us, the stupid human, who didn't know how to use the tool. So give your position to play. Remember when you were five playing with Legos or erector sets or whatever. Give yourself permission to play.
Philip Topham
Once you kind of understand what you can do, think about what it is your business is. And I would really use it to do research about your business, your competitors, so that you can think about what you can do because right now, you have no clue what it can do and what could be possible.
Philip Topham
Once you start looking around, you'll be amazed at how much productivity, different changes you can do. I'm using it for marketing. I'm using it to write presentations. I'm using it to manage calendars. It's amazing what you can do. And every day, it gets better and more amazing. But it's just one step at a time. Just get started.
Niels Brabandt
Excellent. When now people say, "Hey, I think Phil can be really of help to us, maybe as a keynote speaker online or on-site, or maybe he can help us with workshops, coaching, whatever else," how can people get in touch to wrap this interview up? How can they get in touch with you?
Philip Topham
Yeah. The best way is to go to the craftthinking.ai website. There's a contact form there. There's all the information about the book and the speaking opportunities and stuff. Love to help anybody out in the business because I think AI is so remarkable, and it helps. It augments our humanity. It's a co-thinker if we use it as a co-thinker, and we can steer what we want out of it.
Niels Brabandt
Perfect. I think these are the perfect final words. Craft Thinking, a playbook for clear thinking and better decisions with AI. Philip Topham, thank you very much for your time.
Philip Topham
Thank you very much. My pleasure.