Summary

In this episode, I’m joined by Martin Eriksson, founder of Mind the Product and author of The Decision Stack, to talk about how organizations can make better decisions from strategy all the way down to everyday execution.

We talk about the difference between speed and velocity, why empowered teams still need strategic context, and how assumptions exist at every layer of an organization, not just inside individual products.

We also explore what AI changes and, maybe even more importantly, what it doesn’t. Building has become drastically faster and cheaper, but that doesn’t eliminate the need to understand the customer, the business, and whether or not you should be building something in the first place.

If you want to learn more about how strategic choices connect to everyday decisions and how assumptions can be tested throughout your organization you’ll love this episode.

Takeaways

1. Speed without direction is not enough. Teams can build, ship, and test faster than ever, but without strategic alignment they may simply move in different directions more quickly. Martin frames this as the difference between speed, velocity, and organizational momentum.

2. Empowered teams still need strategic context. Autonomy works best when teams understand where the business is going, how it plans to get there, what matters now, and how their work connects to those choices.

3. Assumptions exist at every layer of the organization. They are not limited to products or experiments. Vision, strategy, objectives, opportunities, and execution all rest on assumptions that can become dangerous when they remain unspoken.

4. Strategy is a series of connected choices. The value of a decision stack is not just having vision, strategy, objectives, opportunities, and principles—it is making sure those decisions connect from the top down and can be traced back from everyday work.

5. Use testing and discovery on strategic decisions, not only product decisions. Teams have become good at de-risking what is directly in front of them, but the same tools can be applied to bigger bets around strategy, markets, and the future of the business.

6. Not every assumption can be eliminated, but it should be named. At the strategy and vision level, some choices will always remain bets on the future. Making those assumptions explicit allows leaders and boards to monitor when conditions change and decide whether the risk is still acceptable.

7. AI changes the economics of building, not the fundamentals of product. Faster and cheaper delivery reduces feasibility risk, but teams still need to understand customer value, business value, usability, distribution, pricing, and whether something should be built at all.

Guest Links

Martin’s Website: https://martineriksson.com/

Martin’s LinkedIn: https://www.linkedin.com/in/martineriksson/

Coherent & Wrong: https://www.thedecisionstack.com/coherent-and-wrong/

Summary

In this episode, I’m joined by Teresa Torres. Teresa is an author, speaker and a product discovery coach who I’ve been a fan of for quite a long time.

We chat about why assumptions look different depending on the level you’re working at, why showing your work is so important for alignment, and how using AI inside your courses can give students more opportunities to practice and receive feedback.

We also dig into AI evals - how Teresa uses them to measure the quality of her AI coaching tools, identify failure modes, and systematically improve their performance instead of just trusting that an LLM will get it right on the first try.

If you want to learn more about testing assumptions, designing better learning loops, and using AI without giving up your critical thinking skills, this episode is for you.

Takeaways

1. AI can create a safer space for learning. Teresa found that people are often willing to ask an AI questions they might feel embarrassed asking another person.

2. An AI coach works best when it is grounded in a clear teaching model. Teresa’s interview coach was effective because it was built on years of refined curriculum, rubrics, and explicit feedback criteria.

3. AI can dramatically increase opportunities for deliberate practice. Instead of waiting for an instructor, students can practice repeatedly and receive immediate, personalized feedback.

4. Building AI tools can improve the underlying curriculum. When an agent struggles with ambiguous instructions, it exposes gaps in how the material itself is taught.

5. Evals are simply a way to measure AI quality. Teresa uses evals to identify specific failure modes, track how often they occur, and test whether changes actually improve the agent.

6. Domain expertise still matters. Recognizing that an AI coach has given subtly bad advice often requires deep knowledge of the subject, not just technical skill.

7. Product teams should define what “good” looks like for AI. Generic quality metrics are not enough; teams need acceptance criteria and evals tied to the unique value their product is supposed to create.

8. AI changes the role of the teacher, not just the tools. Teresa is moving away from being the expert who simply delivers answers and toward creating environments where people can practice, get feedback, and learn for themselves.

Guest Links

ProductTalk Website: https://www.producttalk.org/

Teresa’s LinkedIn: https://www.linkedin.com/in/teresatorres/

Summary

In this episode, I’m joined by Barry O’Reilly, he’s an entrepreneur, advisor, author and a friend of mine who works with executives to redesign how their organizations perform. He’s also the co-founder of Nobody Studios, an AI venture studio building and launching over 100 companies.

Barry and I chat about what he’s learned from years of helping corporations experiment, redesign systems, and make better decisions under extreme uncertainty.

We talk about why so many AI transformations start in the wrong place, why leaders should focus on the work before the tools, and how capturing conversations can create a kind of organizational memory that gets smarter over time.

Barry also shares his lessons from building Nobody Studios and how that work influenced his new book Artificial Organizations.

If you want to learn more about testing with AI to redesign how work gets done while making better decisions, you’ll love this episode.

Takeaways

1. Start with the work, not the AI tools. The biggest gains come from redesigning how work and decisions happen, then choosing technology that supports that flow.
2. Human judgment becomes more important, not less. AI can capture, synthesize, and model information at scale, but leaders still need to decide what matters and what action to take.
3. Treat conversations and decisions as data assets. Capturing meetings, transcripts, decisions, and outcomes creates organizational memory that can be reused instead of constantly recreating context.
4. Better systems can outperform experience alone. Deep domain expertise still matters, but rigorous decision-making systems combined with machine intelligence can challenge gut instinct as the default.
5. AI transformations fail when they are treated as tool rollouts. Buying Copilot or another platform does not change how an organization works unless behaviors, processes, and operating systems change with it.
6. Leaders can accelerate adoption by role-modeling experimentation. Admitting “I don’t know,” trying tools in real work, and openly sharing what works and what does not creates permission for others to learn.
7. Start with one decision or workflow. Rather than attempting a company-wide AI transformation, pick a recurring decision or process, test a new way of working, and learn from the result.

Guest Links

Barry’s Website: https://barryoreilly.com/

Barry’s LinkedIn: https://www.linkedin.com/in/barryoreilly/

Artificial Organization: https://artificialorganizations.com/

Nobody Studios: https://nobodystudios.com/