In a previous article we described the Fifth revolution of computing. We finished that by speculating as to the nature of the UI that would come with it. There has to be a new user interface. The reasoning is simple. There are too many new factors involved in the use of mobile devices for any existing user interface to be appropriate. The important factors are:

1. A mobile device can receive alerts direct to the individual.
2. For some devices at least the form factor will be small.
3. Voice is a big factor in mobile devices, and voice input will be a feature because it already is on some mobile phones.
4. Many of the new applications that will drive mobile technology, particularly the use of a mobile device as a remote control device, payment device or as a key, do not require much of a keyboard.
5. The larger user base for mobile devices demands a simple interface that is very easily learned.
6. Profiling will be an element of the interface (recording what a user does and intelligently altering the interface to conform to his/her preferences and common usage).

One only needs to consider a few of these factors to conclude that neither a Windows-type interface or a web browser based interface is going to cut the mustard. Both are too complex and too visually oriented, and neither integrates well with voice.

In the past, user interfaces have not evolved according to a rational set of principles. Instead, an idea has been added to, extended and often damaged as a consequence. For example the Windows type of interface was compromised from the start by its inheritance of function keys and special characters so that experienced users could use short-cuts. These were left in place when pointer based features came along, giving us drop down menus, icons, buttons, slide bars and the rest. In reality what happened was that, fundamentally, several different ways of achieving anything were provided. With interfaces, choice is not a good thing for new users even if it satisfies experienced ones that know what choices to make - and in practice it confuses them too. Similar criticism can be made of the browser based interface.

The possibilities for a new UI might include:

1. The fact that the mobile device will send and receive messages indicates that the interface needs to be message oriented. This suggests an email-like menu of incoming messages (email, voice mail, SMS message, etc.) with sorting, prioritisation, etc.
2. The form factor may be small, but display will have to be an element. The display area of a Palm Pilot is probably OK.
3. Integrated voice will be necessary and this means that the interface must be controllable by voice commands. This suggests a menu type of selection, with limited choices.
4. Voice output will also be necessary.
5. Profiling will be used to diminish choice (as in the current version of Windows) on menus and to organise an individuals use of the mobile technology.
6. Just as all PCs now have to have directory, diary, web access, transactional capability, messaging (i.e. e-mail), the new UI will embody these applications along with spreadsheet capability. In addition however they will also embody the following generic apps:
7. 1. Security of access
   2. Payment capability (linked to a bank)
   3. Remote control (of appliances)
   4. Geographical location (including relevant maps and local directories for restaurants, bars, shops, etc.)

The new UI will be both more sophisticated and simpler than anything that exists right now. However there is a sting in the tail. Whenever a new UI emerges all existing applications have to adapt and this is usually a painful process. It is going to be a very painful process in this instance, because this includes all the web sites that exist out there. To have much of a chance of leveraging the new mobile devices they will have to move into XML and they will have to be able to exploit whatever interface finally emerges.