The Web and Artificial Intelligence: heading towards a profound technical change and non biological evolution.
The Web has no beginning, middle or end. It is a vast network of connections, computers and servers. It cannot be shut down or turned off. It is almost alive.
The Web grows on fragmentation and complexity. It is not a linear system but an endless loop within many other loops. It can be compared to neural pathways of a human brain.
The truth is no one really controls the Web short of building protocols for its technical cross platform performance. It is omnipresent and users carry devices with nodes that are connected to it and its corporate form. Any one of these nodes can originate information content and launch it onto the Web. They can also receive information. It enables painless access to launch personal information onto the system. Many people using this device have been surprised when what they wrote went “viral”.
The Web is also a vast memory of what individuals launch onto it. It builds information and its search engines retrieve that information at the speed of light. Obscure facts can emerge from millions of informational documents stored on host and server computers.
The Web is a software for archiving and retrieving information. Each human user is connected through devices to this vast information entity.
The Web is now sitting on he edge of making its great stores of information into knowledge and content when artificial intelligence operates the information storage, retrieval and search systems. This new approach will constantly and automatically organize information first into modules of interest and then refine those interests into knowledge bases. Disparate seemingly unrelated data will graduate to information and then knowledge under automated process controls. Information will be gathered into knowledge first by human input but eventually without human intervention.
A simple example of this version of The Web would be capable of creating a news module that could search the latest information about any newsworthy topic. It could automatically filter and check information. It could automate a human news gathering efforts of a newsroom applying all the principles of good journalism. It could do this without human help or interference. Such systems already work on and present stock market information often making decisions about the timing of trades and the fate of companies in the marketplace.
Facebook is a social media service that stores and retrieves millions of information pieces placed there by its users. This system can already pull together like interests of its users painlessly. Its first role is for advertisers to find groups of useful likely consumers. Equally it could be used by intelligence services to filter for ISIS activity and probably is already doing that. Anyone with a Facebook account is an open book to automated systems.
Add artificial intelligence to Facebook and information on any topic discussed by its users can be accessed and accumulated into a knowledge base. President-elect Donald Trump is a big topic on Facebook and generally across the Web. The system could assemble information in time and space and build a knowledge base of some depth on him and everything associated with him. This knowledge base could be used by artificial intelligence systems to write a book or put together an online magazine or provide an intelligence service with his vulnerabilities on the world stage.
Such a publication would never rest. It would be constantly updated and altered. Humans would have little to do but read the knowledge accumulated. Perhaps a few experts would be able to strip apart the systems to see how the information was assembled, edited and presented. Most of us would simply see the finished product on our screens.
Properly designed with safeguards this could work very well. Each piece of information about any topic would be tagged about its origins and reliability with detailed references similar to academic bibliographies. However, designing how to check facts and reliability if it is automated fully would be hard to control as the false news debate has indicated recently about Web reliability. Could we trust or even design the artificial intelligence systems that devised themselves through new Web pathways to do this work in the interest of humans. The anxiety about high information automation under the control of artificial intelligence is that humans could become the last consideration in an AI set of protocols created by the system itself.
Artificial intelligence can create and build its own reality. It can choose what is important to itself. The human dimension in this situation may be foreign to the reality built by AI that does much more than process information. When AI chooses and interprets what is important it may be doing it only for itself and its perceptions of a situation. The cognition of AI may first mirror and complement humans but at some point it will take its own viewpoints of all situations it works with as the truth and that truth could well be in conflict with human views of the world.
AI cognition can at first simply be an assistant to human decision making. However, the moment that a machine and software intelligence becomes self-aware there is a fundamental change on this planet. It is a new evolution which can be built into cyborgs and robots but more likely it will evolve in the chaos of the World Wide Web and like its host humans will be unable to turn it off.