I've mentioned in the past the role of pattern recognition changing in the future. Currently, most organizations confine their business intelligence (BI) use to highly skilled analysts. These analysts are expected to have deep industry and process expertise that helps them make sense of what the tool is saying.
In the future, this information will need to be incorporated into a wider range of the organization and packaged so it can be consumed by the target audience. The EDS Fellows have written about the issue of Attention Management and the need to ensure the audience does not become overwhelmed with the information flowing at them. It needs to be provided in context.
There are those who say that wide use of BI will be expensive because training everyone to use the tools and perform analysis will be hard. I believe if it takes that much training it is because the interface was written poorly. Most people do understand the context in which they work. The information needs to be presented in that context (e.g., there is something wrong with milling machine #3) and integrated with workflow. Therefore, little or no training should be required.
Sure, it will be harder in the software analysis and development activities, but at least that is performed by a small group and the activity (if done right) is only done once. BI today is underutilized but by using these kinds of techniques, it will move to be a powerful, continuous use business tool.