Data science
fromMedium
6 days agoHow To Use Andrej Karapthy's Knowledge Wiki Approach for Product Research
Andrej Karpathy's approach to creating knowledge bases enhances AI's ability to remember and focus on data.
Many people in the product community, including myself, actively share their experiences using AI tools to build products. However, far fewer conversations focus on a foundational phase: product research. The quality of product research directly impacts the outcome of the entire design process. As AI tools become increasingly embedded across different phases of the product design process, including the research phase, it's vital to establish a clear, intentional research process that maximizes design efficiency while reducing business risk from poorly informed or incorrect decisions
Shopping online is easier than ever. Remember when you had to juggle 20 tabs, read through endless reviews and buyer's guides, and hunt through promo code sites to find a good deal? In 2025, AI can do all that for you. From researching products and comparing prices to finding deals and even carrying you all the way to checkout, the process is easy and, dare I say, fun.
OpenAI announced shopping research within ChatGPT, "a new experience in ChatGPT that does the research for you to help you find the right products," as OpenAI put it. ChatGPT can now help you research the right product for you. OpenAI wrote, "it asks smart clarifying questions, researches deeply across the internet, reviews quality sources, and builds on ChatGPT's understanding of you from past conversations and your ChatGPT memory to deliver a personalized buyer's guide in minutes."
Let's be brutally honest. Your user personas are probably garbage. They're a collection of stock photos and demographic fluff that you and your team spent a week creating, only to file them away in a Google Drive folder to be forgotten forever. "Marketing Mike," who is 35, married, and "enjoys hiking," isn't helping you make better decisions. He's an art project.