Today, most AI applications run in the cloud on powerful, specialized computers in data centers. But experts say this will not always be the case. There will also be an increasing number of small models that run directly on users’ devices, from PCs to smartphones.
This is called edge AI, after the concept of edge computing. It is meant to signal that the real work is done at the “edge” of the network, not centrally.
This is made possible by advances in chip technology and new methods for developing powerful AI models that require far fewer resources. One approach, for example, is to train a model for a well-defined task rather than as a jack-of-all-trades like ChatGPT.
One advantage is that such locally running AI models do not require internet access and therefore work as usual, even when mobile phone reception is poor. In addition, users’ data remains on their devices and is not sent to the cloud for processing.