One of the topics raised in the Hub4NGI D2.1 deliverable that needs to be addressed by NGI research topics is that of "Decentralisation of Control". A visualisation of the key factors is provided as shown here:
Decentralisation of Control: Key Factors
The report provides evidence from multiple sources including the 2017 DIN Forum Report which acknowledges that the GAFA (Google, Apple, Facebook, Amazon) incumbents are in such a dominant position that it is difficult for European start-ups to compete. The NGI should enable alternative business models and infrastructure to support alternative solutions to the current centralised service offerings.
The Hub4NGI deliverable also cites input on the subject from the Report of the Policy Workshop on Next Generation Internet, Centre for Science and Policy, Cambridge Computer Laboratory from March 2017. Participants in that workshop raised concerns about how the GAFA incumbents were in fact best positioned to exploit emerging AI techniques and "further entrench their market dominance". The workshop also discussed how the principles that were the drivers for the TCP/IP protocols, namely lack of ownership and content agnosticism, no longer reflect the current situation in the Internet. The report of the NGI public consultation published in March 2017 echoes this concern noting that 80% of participants agreed that it was important to "limit the increasing power of dominant platforms aggregating data" and that 74% saw decentralisation as a means of avoiding monopolies. The consultation report also indicates that participants felt strongly that "the Internet should ensure diversity, pluralism and a right to choose".
To what degree does the current ICT-24-2018 call for proposals address these issues? What could future calls also include in this area?
2 comments on "Dencentralisation of Control"
How about ISOC, ICANN, IAB, DNS, TLS certificates, and all that?
I can see that the analysis above addresses mostly economic problems at the applications space. However, is the NGI at all interested in how the underlying infrastructure is administered, including the roles of ISOC, ICANN, IAB, etc. For example, would there be any point to consider how the Domain Name System (DNS), root certificate management in browsers, etc, create centralised or less centralised control points in the network? Or is NGI only interested in things happening in the cloud and social media domain?
Dentralised infrastructure is also of interest
Decentralised infrastructure is certainly also of interest to NGI moving forward.
The above referenced deliverable also has recommendations for research to be undertaken to determine: