Local AI needs to be the norm

A Hacker News discussion argued that local AI should be the default, not cloud-only access.

Commenters said on-device models improve privacy, resilience, and user control by reducing dependence on a few major providers.

The thread also questioned the economics of open-weight releases, noting that training frontier models is extremely expensive and incentives are often strategic.

Examples like Llama, Gemma, DeepSeek, Whisper, and MBP M2 highlighted growing interest in practical local AI adoption.