Software developers face a new wave of LLM evangelists who champion agentic coding as the future of programming. The author, a productivity skeptic, recounts personal frustration with prompt‑driven development, noting that the process demands constant babysitting and often yields buggy code. Despite this, many claim higher output in the industry.

Critics argue that the hype masks underlying insecurity among developers who fear obsolescence. The author suggests that evangelists project their own doubts, framing agentic tools as a threat to traditional skill sets. Yet, for non‑experts, these models unlock creative possibilities that were previously out of reach in the digital era today.

Looking ahead, the debate will hinge on whether LLM tools can reliably replace human judgment or merely augment it. Industry leaders are testing hybrid workflows that blend human oversight with automated code generation. Stakeholders should monitor how these experiments affect productivity, security, and the future skill demands of software teams.