Section 3 / Page 84

The ability to learn, see, interact  in the way described comes back to the simple equation of manipulating data around a system or network. The fact that the data will be manipulated in the way described should be seen as a software engineering challenge rather than a hardware specific problem. The learning ability of some of the modeling agents employed were to be limited to a number of walled garden environments or only certain users. In this way, it should have protected the developers from an uncontrolled growth within the A.I. system. T3 Skynet gets let out of it's box?

The ability of lots of simple networked software mechanisms to transform their behavior and evolve into something more complex, should be something to be watched very closely and controlled, by any developers designing a system like the MNN.

The concepts database was to have the ability to study how different social or ethnic groups behaved, e.g. in Japan, the system would have been more likely to see peoples avatars bowing to each other, rather than shaking hands. The ability to learn concepts in this way, rather than being hardwired or programmed means it should have been capable of changing its behavior to suit the changing moods or trends within our own society. The ability to learn in this way and on the scale envisioned could have given the systems operators access to a real-time system capable of at least emulating human behavioral patterns.

The evolution of user interfaces / VR databases and the concepts / behavioral database, was to have allowed the developers to produce an increasingly smart system for lower and lower cost, with the MNN eventually helping in its own development. This would depend upon how much power we wish to give these A.I. creations. The ability to control A.I. should be seen as a primary concern to anybody who has truly grasped just how intelligent these systems could become, the MNN could have evolved very quickly. A.I. evolution will have to be controlled, because an uncontrolled growth could result in many problems. Section 4 deals with these problems.

People should keep in mind, that the big M is currently making the most money from the man machine interface, (Windows / Office etc) and seeing as how strong A.I., is in essence the ultimate expression of the human machine interface (HMI), then Microsoft is one of the biggest players in the A.I. field, despite what they may say publicly. So the MNN or concepts database, as I like to see it, is just an extension of the philosophy we see incorporated into Windows Vistta etc. Were every time a problem occurs, it sends a report back to Microsoft, so their engineers, can then work on the problem (EMC phone home?). If you extend this philosophy into the near future, then you should be able to see, that  eventually end user input will be used in the way described and eventually by all players, in the man machine interface market. Think of it, as a massive feedback loop, it makes me wonder if this means, all systems will eventually become fault tolerant?

• Predicted Bandwidth: -

Laziness has struck again - most of this was written - pre 2006 - but most of it, is still relevant.

WI-FI the name given to the 802.11b standard, is the standard of choice for wireless communications for most portable net access devices. This system and its derivatives, (there is a whole slew of 802 standards), already allows for data transfer rates of over a 100Mbits/sec. This is still slow when compared with some up and coming technologies, such as UWB ultra-wideband band (a 1000 times more capable than Wi-Fi), this system, may possibly steal the market from 3G / 4G / WI-FI max / Mesh radio etc, click here for a breakdown of wireless terminology.

VDSL (Very High Speed Digital Subscriber Line), is currently seen as the next step in the network bandwidth war, at least to the home and small business user. Bandwidth is always going up as switching equipment /cable and laser technologies improves. IBM's 0C768 hardware was the high end, in network switching equipment, but this has now been superseded by switching equipment operating at over 500 Ghz, allowing bandwidth of well over 100gbs, this is already old hat of course?, as we move ever closer to systems using Nanocomputer Architecture. Terabit networks are already a reality and the move towards Petabit Networks is sure to be in the pipeline. The ultimate goal of the networks, is to get every home user connected via a fibre optic link and every portable platform to utilise a high speed pervasive untethered web access system, which will eventually allow everybody all the bandwidth they could ever possibly need. See chart for terminology and bandwidth breakdown. BT calls it's future plan 21CN (21st century network), basically a plan to lay fibre optic to all UK homes probably by 2015. Although there was a plan to lay fiber optic to every UK home in the 80's and 90's, but the government intervened in it's infinite wisdom?