An NGI dominated by Social Media
Despite or thanks to the progress of the Internet of Things, I think that the next generation Internet will still be dominated by social media. The IoT will enormously increase the amount of data ingested by the Internet, but those data - like the continuous stream of images through our eyes - will inevitably be lost, if not selectively saved by a human consideration of their relevance. Just as only a few amongst innumerable transient images are fixed in memories by our awareness, similarly the web needs to rely on some form of social discussion and consensus to filter and promote the most valuable information. Even when automated data analysis is available, nevertheless the analysis output is just another form of data (aggregated instead of raw), that still needs the human approval to be considered knowledge worth of attention and spreading.
Humans both consume and provide information
Social media have proved to be more effective than more traditional sources at spreading and consolidating knowledge, even in niche technical domains. Today there is a flourishing of social platforms that compete between themselves and try to differentiate e.g. with respect to the managed topics, supported media and content presentation. Sites like Twitter and Facebook are efficient means of information spreading and trends identification, while other sites like StackOverflow, Reddit and Quora are effective means of filtering the most useful information via some sort of up-voting policy. All of them are currently characterized by the fact that both the information consumers and providers are humans. There is also limited interoperability among them: it is possible to share some content from one medium to another, but it is not possible e.g. to define a stream of information that automatically aggregates content from different media or to compose a custom user interface displaying the aggregated content.
Social Media - Evolved Media
I expect that social media will significantly evolve with the automatic ingestion and processing of data coming from the Internet of Things, presenting such data in a rich graphical format, entertaining and helping the human end users (that's why these evolved media should continue to be named social media). We already know examples of aggregated data coming from sensors often in (quasi) real-time, like air pollution, weather forecasts or road traffic conditions, and we can expect more and more examples of sensor data integration into social media: "Hi Jake, I just lost the last bus and I have to wait for you, but please speed up, you see that is cold and raining out here". "Hi Mary, I'm coming as soon as possible, but you see that I'm currently stuck in the traffic". The integration of different digital services is nothing new.
The Human factor as a bridge
We have already seen a long story of achievements and promises about building complex services via the composition of elementary functionalities. These are indeed the principles that inspired software service-oriented architectures or, more recently, software micro-services. Therefore I don't expect giant leaps in machine-to-machine interaction. But the addition of the human factor in bridging and orchestrating the different web services could be revolutionary. In fact, we like to choose the shop where to buy our shoes, the restaurant where to eat our food and would also like to choose our job (unfortunately, that's not always the case): overall we like the freedom to choose the pieces that make up our life. The same freedom should be exercised when choosing and composing the web services that dictate how people use the web and interacts via the web.
Choose your service to suit your profile
The web should resemble a road with many different shops offering different goods in different flavours, and everyone should be allowed to choose and compose which services better suit to his social profile. What seems missing is a platform supporting the user in the selection and integration of multiple social services. Such a platform should ensure service interoperability and composition. Front-end services would assist the user presenting the results of the back-end services' interactions. Special meta-services should support the user in the front-end/back-end service selection and composition process. Other meta-services could be used to classify and rate the many different service offers.
The Human factor defines the pipeline of services
Our imagination should not be limited to the real-time presentation on a smart-phone screen of Jake's car speed and position, together with the estimated time to Mary's bus stop and the weather conditions of her location. The web could be used to carry on what-if analyses that are too complex to be treated only by computational means. The human factor could manually define and orchestrate a pipeline of different services, in order to correlate data streams from different sources and produce heuristic results.
There is the need of a platform for facilitating the aggregation, the selection, the review and rating of useful information coming in different formats from multiple sources, a platform open to information providers and consumers and scalable with respect to the ever increasing amount of input data. A platform supporting service interoperability. A platform supporting meta-services dedicated to service classification, selection, composition and rating. A platform supporting services to pipeline data streams from disparate sources to carry on what-if simulations.
The role of the public regulator
Such a unique versatile platform does not exist and will never exist as a service from a single vendor, but it could be the result of the interaction via web of multiple data streams from different providers, processed and aggregated by multiple services from different vendors. The next generation Internet could be that platform. The public regulator will play the key role to support the definition of the web standards and the implementation of the web infrastructure necessary to realize and evolve the web platform. The public regulator will play the difficult role to guarantee the open interconnection to the web platform for information and service providers. The public regulator will play the challenging role to verify the interoperability of the provided services, in compliance with the emerging standards and practices.