Until
next week,AI and The Future of the Internet: Smarter
Dateline: November 2, 1997
This is part 3 of a 3-part series about the Internet, AI, and their mutual significance. In the first article, we sketched out ways in which the Internet has developed, from various perspectives, for the better. In the second article, we looked at the growth in the size of the Internet which its betterment fostered. In this final article on the subject, we see how growth in size (and attendant complexity) fosters the development of "smartness" or intelligence within the Internet.
Complexity (chaos) theory and our own experience tell us that as things grow bigger they also tend to grow less stable. Any network manager can attest to that, and even the mighty Microsoft’s Windows NT remains humbled by UNIX’s ability to outperform it in handling the complexity of large-scale networks without falling apart.
But even UNIX has its limits, and increasingly network managers are turning to AI-based network management solutions to make sure the mail gets through. The ultimate goal is for networks to maintain, diagnose, and to some extent repair themselves, automatically adapting to component failures, overloads, equipment changes, removals, and additions, and only asking for human help to insert/remove plug-and-play modules when necessary.
Network users, again as any network manager or software engineer will tell you, are an even bigger pain in the butt than the network itself. They demand that the Internet itself takes on the burdens of finding relevant information and filtering unwanted information from the growing and bewildering confusion of Internet services and information bases. Their needs are increasingly being met by AI-based tools such as searchbots and spam/porn/hate-group filters.
Non-users who could afford and find access to the Internet but are daunted by keyboards and the need to type will soon find that barrier removed, as ASR (automatic speech recognition) software becomes more powerful than today’s version 1.0’s. Even the mouse is at best an awkward, intrusive, and imprecise device, destined for the scrapheap of history when data gloves and suits become better and cheaper and VR becomes the standard interface.
Finally, the language barrier, which makes the mainly English-language Web at best a limited and at worst a useless resource for about three-fourth’s of the world’s population, will be smashed by AI-based ALT (automatic language translation) software that will not only handle written materials (presenting English-language Web pages and email messages in the user’s choice of language) but also spoken language, sitting among participants in international VR conferences and providing simultaneous multi-way oral interpretation.
Provided there are no major problems with the Internet itself, all of these wonders and more shall come to pass. However, a major problem facing the Internet as it grows smarter is network security. I don’t mean the security of transactions over the Net, but of the Net itself. As we discussed in my article about Bots, bots (and viruses) created by malignant or merely careless programmers can wreak havoc. If they can get to the routing tables and mess them up, for example, they would bring the Internet to a halt. There have been some serious incidents already, and as the following quote shows, the U.S. government, for one, is deeply concerned:
U.S. IN DANGER OF CATASTROPHIC CYBER ATTACK
Robert Marsh, the chairman of the Commission on Critical Infrastructure Protection says that neither government nor industry now has the means to protect the United States against computer attacks that could shut down communications and power grid. The commission will deliver its report to President Clinton this week.
Montreal Gazette, October 8, 1997
But think for a moment. An equally valid headline might read:
CALIFORNIA IN DANGER OF CATASTROPHIC EARTHQUAKE
The devastation caused by such an event is unquestioned, but in the grand scheme of things it will not be total, irreversible, and everlasting. On the contrary: it will be partial, repairable, and temporary. It will also lead to safer structures, better emergency and rescue drills, better prediction methods, and less devastation next time around. I maintain that nothing short of a massive asteroid strike, or God, can shut down either California or the Internet.
We can expect, therefore, that the developments I have discussed in this and the previous two articles will continue to the point where there are billions of people communicating in their own native tongues over terabit networks with other people and with intelligent machines. And life will start to get really interesting.
Until
next week,
NEXT WEEK: Tissue Engineering: Part of android evolution.