AI coding assistants offer speed, but what impact do they have on quality, trust, and the art of architecture? Shweta Vohra and Grady Booch delve into a thoughtful perspective on how architecture needs to adapt as machines start coding alongside humans in the first episode of Next Gen Architecture Playbook. The software industry is currently experiencing its third era of prosperity, with AI being a part of it rather than the root cause. Progress has shifted from algorithms to objects to platforms and globally distributed systems. AI tools now exist within this platform era as an additional layer of abstraction, speeding up how architects construct and integrate systems that are not entirely under their control. Despite automated design processes, architecture still involves making decisions that are costly to alter. AI can swiftly produce designs, patterns, and code, but architecture remains the collection of crucial choices that impact long-term expenses, resilience, and development. AI enhances efficiency by elevating abstraction levels, not by replacing cognitive functions. Similar to how compilers operated in the past, LLMs delegate low-level tasks to machines, enabling humans to focus on higher conceptual aspects. The primary concerns include skill depletion, standardization, and overconfidence. LLMs often guide teams towards common practices and may subtly diminish learning opportunities for junior engineers. Human responsibility is essential in AI-assisted engineering. Even when AI generates or assesses code, accountability for quality, safety, and results rests with human architects and engineers. All architecture involves design, but not all design constitutes architecture. Architecture embodies the critical design choices that define a system’s structure and function; significance is determined by the cost of modifications required. Creativity plays a vital role in the design and architecture process. While AI can aid in creating designs, creativity remains a human attribute rather than an aspect of AI capability. AI tools are simply instruments that should not supplant the human factor in software architecture. Architects must remain actively involved and accountable for their decisions. Embracing AI in software development necessitates caution regarding potential skill erosion among new engineers. Ultimately, architects must steer the use of AI tools while retaining a firm grasp of architectural fundamentals and human ingenuity.