Are today's most advanced AI models really capable of “thinking”? Or are we simply projecting human-like reasoning onto machines that are fundamentally limited in how they solve complex problems? In this episode of the Professor Insight Podcast, we dive into a provocative new paper from Apple titled The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models. It explores how some of the most powerful reasoning models — like Claude 3.7 Sonnet Thinking, Gemini Thinking, and OpenAI's o1 and o3 — struggle when problems get even modestly more complex.

The researchers tested these models on classic puzzle environments like Tower of Hanoi, River Crossing, and Blocks World — environments that allow precise measurement of reasoning complexity. The findings are surprising: despite their promise, these models hit a “reasoning wall.” They collapse in accuracy as complexity grows, underutilise their available thinking capacity, and even “overthink” simple problems. Apple identifies three distinct regimes where these models either outperform, flounder, or completely fail — and the implications are significant.

But the paper hasn't landed without controversy. Critics argue Apple’s conclusions are overstated and possibly self-serving, especially as the company faces pressure over lagging behind in AI development. Is this research a serious warning about the current limits of reasoning in AI? Or is it a carefully timed narrative to reshape public expectations? Tune in as we unpack the science, the backlash, and the broader debate on what it really means for AI to “think.”

If you’ve listened to Episodes Five and Six of the Professor Insight Podcast, you’ll know we’ve already laid the groundwork on what Agentic AI is and why it matters. But this week, we’re taking it a step further. This episode is your practical guide to building AI agents—an extension of our earlier discussions, now grounded in real-world application. Based on OpenAI’s newly released Practical Guide to Building Agents, we distill a technical framework into actionable insights tailored for business leaders, strategists, and product teams.

We explore what it actually takes to develop your first AI agent—from choosing the right use case and designing safe, scalable workflows, to configuring models, tools, and instructions that help agents operate autonomously and intelligently. This isn’t just about writing prompts. It’s about building systems that make decisions, take action across platforms, and adapt in real time—all while staying aligned with business goals and compliance requirements.

Whether you're trying to reduce operational friction, tackle high-complexity workflows, or enable more intelligent automation inside your organisation, this episode will give you the clarity and confidence to start building. With insights on orchestration, human-in-the-loop design, and guardrails for safety and governance, this is your field guide to the next evolution of AI in business.

In this episode of the Professor Insight Podcast, we explore a fascinating and timely question: how is generative artificial intelligence beginning to reshape academic research itself? Drawing on a newly published academic paper from the Journal of the Academy of Marketing Science (April 2025), we examine how large multimodal models like ChatGPT-4o could revolutionise the way marketing scholars generate ideas, build theories, design experiments, and analyse data. Authored by Kiwoong Yoo, Michael Haenlein, and Kelly Hewett, the paper—titled “A whole new world, a new fantastic point of view”—offers a serious look at how generative AI might soon become a research partner, not just a productivity tool.

Rather than speculating, the authors put AI to the test by replicating the entire research process of 35 published consumer research articles using ChatGPT-4o. They applied advanced prompting strategies like chain-of-thought prompting and carefully evaluated the AI’s performance across key stages such as theory development, pilot testing, and data analysis. What emerged is a nuanced picture of where AI excels—like generating conceptual frameworks and simulating study designs—and where it still struggles, such as interpreting human behaviour or conducting robust statistical analyses.

This episode is a must-listen for anyone curious about the intersection of AI and knowledge creation. Whether you are a researcher, educator, PhD student, or simply someone fascinated by how ideas are shaped and shared, this conversation offers a rare look at how AI may redefine the boundaries of academic research. We explore the promises, the limitations, and the ethical questions that come with using AI as a co-investigator in scholarly work.

In this second instalment of our two-part series on prompt engineering, we take things to the next level. If you have ever wondered how to get large language models to think more deeply, reason more clearly, or act more independently, this episode is for you. We explore advanced prompting techniques that unlock far more than simple question-and-answer interactions—techniques that help AI reason, plan, and execute in more intelligent ways.

For more technical users—those building with AI, developing workflows, or exploring the limits of what generative AI can do—we will be covering Chain of Thought prompting, which helps guide models through complex reasoning steps, and Tree of Thought prompting, which enables the model to explore multiple problem-solving paths in parallel. We will also briefly touch on ReAct prompting, which allows models to not just reason but take action using tools and external resources.

Even if you’re not deeply technical, this episode still has plenty for you. We cover Automatic Prompt Engineering, a technique where AI helps refine its own instructions for better performance, and explore how to write powerful prompts for code generation, translation, explanation, and debugging—even if you're not a developer.

While some of the strategies we discuss go beyond what most non-technical users will use day to day, they offer a deeper look into what is possible when AI is used with precision and purpose. And to wrap it all up, we share a comprehensive set of best practices in prompt engineering that will help you write clearer, more effective prompts regardless of your background or use case.

Prompt engineering is quickly becoming one of the most valuable skills in the AI era, yet most people still treat it as trial and error. In this first episode of our special two-part series based on the Prompt Engineering resource authored by Lee Boonstra at Google, we break down what prompt engineering actually is, why it matters, and how understanding the mechanics behind large language models can dramatically improve the quality of your AI interactions. From token prediction to temperature settings, we explain how generative AI decides what to say next—and how you can guide it with more precision.

We cover everything from basic to intermediate techniques, including single-shot and few-shot prompting, system and role prompts, and often-overlooked features like output configuration and randomness controls. These are the tools that separate casual users from those getting truly valuable results. You will learn how to write clearer prompts, control creativity, and design better outputs for everything from content generation to complex decision support.

Whether you're a student, professional, or a business leader, this episode will give you a strong foundation in how to think about prompting. It is practical, actionable, and sets the stage for Part Two, where we will explore advanced strategies like Chain of Thought, Tree of Thought, and more. If you want to work smarter with AI, start right here.

In this episode of the Professor Insight Podcast, we break down one of the most in-demand skills in today’s AI-powered world: prompt engineering. With more companies actively hiring prompt engineers and more people using tools like ChatGPT, Gemini, and Claude in their daily workflows, knowing how to craft effective prompts is no longer optional — it is essential. Drawing from my own deep dive into research papers, online courses, and hands-on experimentation, I introduce a practical, proven approach to prompt design: Google’s Five-Step Prompting Framework.

We explore each step in this accessible model — Task, Context, References, Evaluate, and Iterate — and I share real-world examples to help you apply these concepts right away. Whether you are struggling to get useful responses from your AI tools or simply want to improve how you interact with language models, this episode provides a structured foundation for writing better, clearer, and more effective prompts. We also touch on helpful techniques like using personas, setting output formats, and refining your instructions when things don’t quite work as expected.

This is your go-to guide for Prompt Engineering 101, and it is perfect for beginners or anyone looking to sharpen their skills. But that’s not all — as I put this episode together, I realised there was still so much more to explore. That is why I’m launching a follow-up two-part special series focused on advanced prompt engineering, where we will cover more technical strategies for guiding language models to reason, reflect, and act with greater sophistication.

Update (11 May 2025): Audio issues after 3min have been fixed. Thanks for your patience, and apologies for the inconvenience.

In this episode of the Professor Insight Podcast, we take a closer look at one of the most talked-about developments in artificial intelligence: Agentic AI. After our last episode introduced the concept, many listeners asked for a deeper dive. So today, we are breaking down what Agentic AI actually is, how it works, and how it differs from generative AI, large language models, and traditional AI workflows. If you have been hearing the term but are not quite sure what it means in practice, this episode will give you the clarity you need.

We explore the defining traits of Agentic AI, including autonomous decision-making, goal-driven behaviour, real-time planning, and tool use. Unlike models that simply respond to prompts or follow fixed sequences, Agentic AI is capable of taking initiative and solving complex problems independently. We also examine how this compares to generative AI, which is focused on content creation, and AI workflows, which rely on predefined steps and human direction.

So why should business leaders care? Because understanding these distinctions is critical to making smart choices about automation, innovation, and competitive strategy. Agentic AI opens up powerful new possibilities for efficiency, scalability, and intelligent decision-making across industries. Tune in to find out how this next wave of AI could reshape the way your organisation works and delivers value.

In this episode, we plunge into the groundbreaking realm of Agentic AI. These aren’t just smart assistants—they’re autonomous agents capable of making decisions, taking action, and learning on the fly with minimal human input. Agentic AI is no longer a distant future – it's happening now. Businesses across the globe are rapidly exploring and integrating this technology to gain a crucial competitive edge. Tune in to discover how these sophisticated systems are moving beyond passive assistance to actively make decisions and execute tasks with limited or no human intervention. This podcast is based on the “Mastering AI Agents” ebook resource by Galileo, and the “Agentic AI - the new frontier in GenAI” report by PricewaterhouseCoopers.