Перейти к содержимому

How a Software Engineer Actually Prompts AI

nitty gritty English

0:00 / 0:00

How a Software Engineer Actually Prompts AI

252 просмотра · 12 дней назад
nitty gritty English
111 подписчиков
252 просмотра · 12 дней назад
Software architect Aravind Reddy joined me to explain how language models actually work and why your prompts are probably failing. Whether you’re learning English and prepping for a high-stakes client call, or simply tired of getting robotic, useless answers, this is the practical, grounded masterclass in making AI work for you. (And yes: we officially believe in Aravind :). Aravind's prompt generator: https://promptgateways.com TIMESTAMPS 0:00 The hard truth: AI doesn't have a brain 0:33 Meet Aravind: software architect, problem solver, and absolute sweetheart 1:38 Next-word prediction (or: why it just guesses) 4:05 How transformer models read the whole room at once 5:30 The safe deposit box test: why context changes everything 7:40 The foundation: Role, Task, and Context 10:25 Stop talking to AI like a search engine 14:25 Why AI hallucinations happen (and how to minimize them) 15:02 Advanced tier: Loop engineering and tool harnesses 19:41 Live Demo: Prompt Gateways in action 22:05 Roleplaying professional English for real meetings RESOURCES MENTIONED Explore Aravind’s prompting tool: Prompt Gateways