In this episode of The Naked Product, I sit down with Harsh Joshi to peel back the layers of AI Engineering, a field often shrouded in mystery yet central to how generative AI is reshaping products and society.We start with Harsh’s personal journey into AI, before breaking down what “AI engineering” really means in simple, human terms. From there, we dive into the heart of the discipline: how it differs from data science and ML research, and why the unglamorous parts often matter the most in making AI products actually work.The conversation then unpacks Large Language Models (LLMs) vs. Reasoning Models:What makes reasoning models like DeepSeek R1 different from GPTs and Claudes?Is reasoning in AI just probability on steroids, or are we inching closer to structured thinking?What would surprise a non-engineer if they could peek inside?Shifting into the product lens, we explore:The common mistakes PMs make when building AI-first productsHow to balance latency, hallucinations, and costs against user needsWhat actually moves the needle in fine-tuning, RAG, and deploymentFinally, we zoom out to the future of AI engineering, from neurosymbolic AI to whether every software engineer will need AI engineering skills, and close with a reflective discussion on whether AI is learning to truly understand the world, or just to mimic it infinitely.This episode is a blend of personal stories, deep technical insight, and big-picture reflection, a Lex Fridman, style conversation designed to leave you both informed and thoughtful about where AI is headed.👉 Perfect for: AI enthusiasts, product managers, engineers, and anyone curious about the real work behind today’s most powerful AI systems.

Ver todos los episodios