The short answer: on-device AI does not conveniently fit the narrative of Wall Street and VCs expecting massive returns from cloud AI.
It is no secret that the devices we carry around with us have built-in processing to support locally installed AI models and execute AI-centric tasks.
In fact, AI can even live in a web browser. Take Google’s Gemini Nano, for example.
Tasks we commonly associate with AI, such as writing, editing, data extraction, data sorting, reformatting, and much more can readily be handed on our devices, without necessitating calls to ChatGPT or Claude which live in datacenters. Some coding-related activities can be accommodated with local AI as well.
I would go even further, and posit that the bulk of our everyday back-office and front-office work can be automated, streamlined, or even replaced by AI running on our computing devices. On-device AI is the great denial that tech investors conveniently ignore.
Don’t forget that like anything AI, models running on our devices are only going to get better over time, with more capabilities and practical use cases to follow.
AI on our devices means energy savings, good vibes for the environment, and no compromise to our data privacy.
But it also looks to be a threat to the big cloud AI companies. Henceforth the unwavering ignorance on Wall Street.