I'm done with LLM-through-chat-experience (www.thoughtfultechnologist.com)

In a reflective piece, a tech enthusiast shares their disillusionment with the limited effectiveness of conversational experiences with Large Language Models (LLMs). They express concern that while LLMs are marketed as tools for enhancing intellectual work, they often fail to facilitate genuine understanding, likening interactions with AI to a misleading emotional connection. The author highlights the inherent gaps in communication, where language can never fully convey meaning, and emphasizes that LLMs, lacking independent understanding, can only mimic human-like interactions without the depth that comes from lived experiences. This sentiment echoes ongoing debates in the AI/ML community about the potential downsides of over-relying on AI for intellectual tasks. The author raises significant technical implications, suggesting that without exposing LLMs to a broader context of human experience, they remain mere generators of tokens rather than true conversational partners. The piece sheds light on the “automation paradox,” advocating for a balanced approach where certain intellectual tasks are deliberately retained to nurture critical thinking and judgment—skills that could be diluted in a heavily automated environment. The call for mindful engagement with AI tools underscores the need for a nuanced understanding of their capabilities and limitations in human communication.