The latest articles on DEV Community by Tushar Vashishth (@tushar_vashishth_45ef7ac3). https://dev.to/tushar_vashishth_45ef7ac3 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4084832%2F63af62a7-b899-4121-b179-18bfc66a660c.png https://dev.to/tushar_vashishth_45ef7ac3 en Tushar Vashishth Wed, 19 Aug 2026 13:33:15 +0000 https://dev.to/tushar_vashishth_45ef7ac3/were-getting-really-good-at-adding-ai-to-things-8ii https://dev.to/tushar_vashishth_45ef7ac3/were-getting-really-good-at-adding-ai-to-things-8ii <p>Add an LLM here.<br> Add a copilot there.<br> Put an agent in the workflow.</p> <p>Done. “AI-native.” 😅</p> <p>But building <em>around</em> AI feels like a different problem.</p> <p>Where does the right information come from?<br> Which model should handle the job?<br> What should an agent actually do?<br> And probably the biggest one — how do we know the result is good?</p> <p>I've been looking at an idea from AiDOOS around this called <strong>RAMP — Retrieve, Agents, Models, Proof</strong>.</p> <p>Still early, but I think there's something worth exploring here.</p> <p>Maybe being AI-native is less about adding AI...</p> <p>and more about <strong>rethinking how the whole system works</strong>.</p> ai opensource python webdev