My latest for KnowTechie is a criticism of “grief tech,” the growing number of AI chatbots that are trained with data of deceased loved ones. These tools, almost all seem as predatory as a Victorian-era séance.
AI’s have simple objectives of satisfying a user. Large language models are prone to sycophancy and sandbagging. AIs will answer subjective questions, flattering their users’ stated beliefs and endorsing common misconceptions when users appear uneducated.