The fact that we're slowly removing the apprenticeship layer and passing it off as "productivity gains" should be far more alarming. Just imagine. The kind of work that is best suited for AI, such as boilerplate code, simple CRUD endpoints, and basic component wiring, is actually similar to the work that helps in training junior developers. It was not that we were working on those tasks because th
Applicant Tracking Systems used to be boring. For most of the 2010s, an ATS was essentially a database with a careers page bolted on top: a place to dump resumes, push them through a pipeline of stages, and email rejections in bulk. The interesting work happened around it, not inside it. That has shifted in the last two years, and the shift is deeper than the marketing pages suggest. I have spent
When you have 5 unrelated questions, should you pack them into one message to the LLM, or send 5 requests simultaneously? Which is faster? Splitting into multiple independent parallel requests is almost always faster. This isn't a gut feeling — it's determined by the underlying inference mechanism of LLMs. Let's walk through the reasoning from first principles. To understand this problem, you firs