In this episode, we dive into the 2025 State of Engineering report to unpack a startling contradiction: while 92% of developers have adopted AI coding tools, overall system throughput and stability are actually declining. We explore the "almost right" crisis, where the time saved in drafting code is being swallowed by the "Verification Tax"—the immense cognitive load required to debug subtle, AI-generated errors.

We analyze the MR2 randomized control trial, which revealed that developers using these tools took 19% longer to complete tasks despite predicting they would be faster. The discussion also highlights a widening talent gap, as senior developers use "architectural intuition" to ship 2.5 times more AI-generated code than juniors, who often struggle to audit hallucinated outputs.

Key topics discussed include:

• The Verification Tax: Why hunting for subtle AI bugs consumes the time gained during the drafting phase.

• The Talent Bifurcation: How seniors act as force multipliers while juniors face a "hollowed-out career ladder".

• Redefining Metrics: Why the industry is moving away from gaming-prone metrics (like the McKinsey framework) toward DevEx (Developer Experience), DORA, and SPACE.

• Vertical AI vs. Wrappers: The collapse of generic AI wrappers in favor of vertical AI SaaS that owns the entire workflow.

• The Apprenticeship Crisis: A provocative look at how we build the next generation of senior engineers if AI automates the very tasks that build intuition.

This episode is a must-listen for engineering leaders and developers trying to find the balance between aggressive automation and maintaining high-quality, sustainable systems in 2026.

In this episode, we explore what some are calling "The Great Correction"—a monumental shift in software architecture away from the "scale at any cost" mentality of the last decade. We discuss why industry leaders are abandoning the complexity of microservices and the public cloud in favor of efficiency, cost control, and sovereign infrastructure.

We break down high-profile case studies, starting with Amazon Prime Video's high-stakes move back to a monolith, which resulted in a 90% reduction in infrastructure costs. We also dive into the "10-million-dollar question" of cloud economics, examining how 37signals saved millions by exiting the cloud and moving to self-owned hardware.

Key topics discussed include:

• The Microservices "Hangover": Why logical modularity doesn't always require physical separation and the rise of the "Modular Monolith".

• Infrastructure Repatriation: An analysis of the "Egress Fee Trap" and how companies are regaining financial sovereignty by owning their own servers.

• Systems-Level Optimization: Why Discord transitioned from Go to Rust to eliminate "garbage collection" lag and achieve absolute performance stability.

• The Rise of the "One-Person Unicorn": How AI convergence and serverless computing are allowing single founders to build what once required massive engineering teams.

• Performance per Dollar: Why the new era of software design is defined by precision, sovereignty, and simplicity over blind growth.

Whether you are an engineering leader managing rising cloud bills or a developer interested in high-performance systems, this episode offers a roadmap for the architectural shift defining 2026.

In this episode, we dive deep into the 2025 Cross-Platform Report, exploring how the two giants of mobile development—React Native and Flutter—have fundamentally rewritten their cores to move past the performance debates of the past.

We break down the technical shift from "performance gaps" to "capability gaps," examining how React Native’s new JSI (JavaScript Interface) and Fabric renderer have eliminated the "bridge" bottleneck to allow for seamless, synchronous communication with native code. On the flip side, we look at Flutter’s Impeller engine, which uses Ahead-of-Time (AOT) compilation to eliminate animation "jank" and provide pixel-perfect control.