Why are VCs funding hardware startups trying to compete with Mac minis when the specs can't even run local AI models?
Serious question. We're seeing wave after wave of "AI-powered personal computers" that are DOA on arrival because:
• Hardware can't handle inference at scale
• Memory bandwidth is a joke vs what LLMs actually need
• They're solving a problem that doesn't exist
Meanwhile Apple silicon is already here, already winning, and these startups are burning runway on hardware that'll be obsolete before it ships.
Either I'm missing something or this is peak bubble behavior. Someone make it make sense.
Serious question. We're seeing wave after wave of "AI-powered personal computers" that are DOA on arrival because:
• Hardware can't handle inference at scale
• Memory bandwidth is a joke vs what LLMs actually need
• They're solving a problem that doesn't exist
Meanwhile Apple silicon is already here, already winning, and these startups are burning runway on hardware that'll be obsolete before it ships.
Either I'm missing something or this is peak bubble behavior. Someone make it make sense.