A veteran software developer's reaction to AI coding tools prompts reflection on where LLMs will land after the current hype cycle crashes. Drawing on the Gartner hype cycle framework, the author predicts most chatbot companies won't survive, but two durable use cases emerge: local LLMs for enterprise intranet search (avoiding data privacy risks), and embedded LLM chips in consumer appliances. The author argues that as AI becomes ubiquitous and unremarkable, human-written content will gain renewed value, and the most lasting AI applications will be local, low-cost, and task-specific rather than cloud-dependent general-purpose chatbots.
Sort: