6G is coming along behind the hype of 5G and unless there is a coherent list of AI activities with a return on investment attached, it is going to be harder for all parties (telcos and vendors) to make investment decisions.
We spent some time looking at projects discussed as developing AI capabilites for 6G and discovered that there was a mix of focus on 6G:
Out of the 130 projects looked at:
Not AI 20% of the projects were not really “AI” or machine learning but really rules-based decisioning. Very useful and the first potential projects to be considered because they are easier to deploy. But not really AI.
Not 6G Specific The majority of the projects were likely to be required to support 5G Advanced and will probably be developed by the time that 6G networks are rolled out. So, technically, not 6G projects.
6G Specific A good number of the projects look like 6G projects and will be developed to support particular features or use cases that are definitely 6G. Looking at these projects, some will be solving more simple problems whilst a minority look set to be providing more – sophisticated intelligence and are probably not the first projects we will see in 2030.
AI 4 AI This last category collects together all AI projects that are needed to support AI as it is added into the network. These are related to requirements such as data management or more sophisticated management of algorithms in the network.
So, is 6G really going to be “AI-Native”?
It would seem more correct to say that 6G will see AI projects added as a continuum to the intelligence and automation inthe network. This will be needed to meet a variety of new challenges from some of the proposed 6G concepts and technologies – and to manage the increasing complexity creep.
Is the phrase therefore just marketing spin? To some extent, yes. And yet it is also a rallying cry for those in the telco ecosystem to create a strong strategy and culture around AI – which is not an insignificant challenge.