The 5G network should be different than prior mobile generations for many reasons. It should become the first generation of mobile networks to enable use by machines, sensors and servers, rather than humans. Faster and lower-latency mobile broadband will be a feature humans will notice, to be sure.
But the primary new categories of usage is expected to be by sensor networks of various types whose requirements are not so much bandwidth as low latency or low cost.
Also, 5G should be the first mobile network that integrates multiple networks, organically, for access (licensed and unlicensed; owned and third party assets). In other words, 5G will be more virtualized than prior generations.
So 5G also should prove disruptive to participants in other parts of the ecosystem, as well as to new ecosystems. That arguably has been the case in the past.
The first generation of mobile created alternatives to fixed phone lines, triggering the huge mobile substitution trend that has decimated use of fixed voice services.
The 2G network created the text messaging business and also demolished the paging business.
The 3G network enabled mobile email and then mobile internet. Most likely would agree that 4G enabled tethering of PCs and other devices, plus video consumption and smartphones with pleasant user experience when accessing cloud data.
But 4G also largely displaced MP3 players and pocket cameras.
It would be unusual indeed if the coming 5G network did not cannibalize some existing communications functions and roles; industry segments or participants.