Most early stage founders think their job is to grow. That belief quietly kills more startups than bad ideas ever will.

In this episode, Thomas Bustos challenges the default startup narrative and argues that early stage founders should delay optimization, ignore premature scale, and focus obsessively on their first five customers. Not leads. Not users. Paying, real ICP customers.

He dismantles the obsession with metrics that don’t matter yet and explains why clarity, learning velocity, and decision quality outperform hustle and headcount in the early days. The discussion explores why premature scaling leads to fragile systems, how visibility enables better decisions, and why early-stage teams must prioritize shared context over efficiency.

Thomas also explains how tracking inputs and outputs creates feedback loops that prevent founders from optimizing the wrong things.

This episode serves as a systems-level analysis of what actually matters for early stage founders before scale, funding, or hiring accelerates complexity.

Top Takeaways:

• Early stage founders should optimize for learning over growth.

• Focus on inputs before outputs to ensure effective decision-making.

• Create a shared map of reality to align the team.

• User feedback is crucial for refining the mission and ideal customer profile.

• Metrics should reflect reality, not be treated as goals.

• High signal users provide valuable insights for progress.

• Building foundations is essential before scaling a startup.

• Learning velocity and decision quality are key to success.

• The early stage is characterized by messiness and uncertainty.

• Navigating through a dynamic map is vital for achieving product-market fit.

Connect with Thomas Bustos

• Thomas Bustos on LinkedIn - https://www.linkedin.com/in/thomasbustos/

• Let’s Talk AI - https://thomasbustos.substack.com/

• Let’s Talk AI on YouTube - https://www.youtube.com/@lets-talk-ai

• Let’s Talk AI on Spotify - https://open.spotify.com/show/6mVjFvdEkZDCTXpIuuSLAP

Hosted on Ausha. See ausha.co/privacy-policy for more information.

In this episode of Let’s Talk AI, Max Berthelot takes us into the heart of his founder journey. It is one that spans B2B SaaS, Y Combinator’s high-pressure halls, and now the health-tech frontier with Lucis.

What starts as a leap from a stable career becomes a story about obsession, discipline, and the uncomfortable truths behind real startup growth. Max shares the moment he realized community was the key to expansion. He talks about the nights spent rethinking distribution, the emotional rollercoaster of working with both technical and non-technical co-founders, and the mindset that transformed Lucis from an idea into a fast-emerging longevity platform.

This is a story of a founder recounting the lessons that scarred him, shaped him, and ultimately strengthened his ability to scale a startup in one of the toughest markets.

If you want startup growth lessons that go beyond clichés, this is the episode to hear.

Top Takeaways:

• Max is the founder of Lucis, focusing on health and longevity.

• Building a community around health can drive engagement and growth.

• Distribution should be prioritized from day one in product development.

• It's important to start with an idea and iterate rather than waiting for perfection.

• Y Combinator provided invaluable lessons and a life-changing experience for Max and his co-founders.

• Relentlessness is key to building a successful startup.

• Continuous learning is essential for founders as they navigate different stages of growth.

• The mission of Lucis is to make preventative health accessible in Europe.

• Max's personal experience with health data drives the mission of Lucis

Connect with Max Berthelot

• Max Berthelot on LinkedIn - https://www.linkedin.com/in/maximeberthelot/

• Instagram: @lucislife_en

Connect with Thomas Bustos

• Thomas Bustos on LinkedIn - https://www.linkedin.com/in/thomasbustos/

• Let’s Talk AI - https://thomasbustos.substack.com/

• Let’s Talk AI on YouTube - https://www.youtube.com/@lets-talk-ai

• Let’s Talk AI on Spotify - https://open.spotify.com/show/6mVjFvdEkZDCTXpIuuSLAP

Hosted on Ausha. See ausha.co/privacy-policy for more information.

The IDE has always been the developer’s command center. But as AI systems mature, the IDE is becoming something far more sophisticated — a hybrid of automation, intelligence, and adaptability.

In this episode, Thomas Bustos and Richard Wang (Clad Labs) dissect the architectural and strategic implications of building AI-powered IDEs. They discuss how real-time feedback loops, adaptive code assistance, and contextual learning models are reshaping both the developer experience and the business of software.

This is a must-listen for engineers, founders, and product builders who want to understand the competitive edge behind AI-driven development tools, and why IDE innovation is the next major battleground for developer productivity.

Top Insights:

• AI is revolutionizing the coding process for developers.

• Integrating ads into IDEs can subsidize development costs.

• User feedback is crucial for product iteration and improvement.

• The future of coding involves AI-native tools that enhance productivity.

• Democratizing access to AI tools is essential for global developers.

• Speed of development should focus on user-centric design.

• Understanding user workflows is key to building effective tools.

• The competitive landscape of IDEs is evolving with new approaches.

• Maintaining a balance between speed and quality is vital in development.

• The mission is to unlock coding opportunities for everyone.

Connect with Richard Wang

• Richard Wang on LinkedIn - https://www.linkedin.com/in/richard-wang-39554994/

• Clad Labs AI - https://www.cladlabs.ai/

Connect with Thomas Bustos

• Thomas Bustos on LinkedIn - https://www.linkedin.com/in/thomasbustos/

• Let’s Talk AI - https://thomasbustos.substack.com/

• Let’s Talk AI on YouTube - https://www.youtube.com/@lets-talk-ai

• Let’s Talk AI on Spotify - https://open.spotify.com/show/6mVjFvdEkZDCTXpIuuSLAP

Hosted on Ausha. See ausha.co/privacy-policy for more information.