Читать книгу Artificial Intelligence for Asset Management and Investment - Al Naqvi - Страница 43

Based upon the above discussion, we are now able to suggest how to architect competitive advantage for our firms. As we digest the above discussion, we can zero in on four core determinants of competitive advantage. These four determinants are the underlying engines that drive value for us and our clients. These are the technological constructs that we need to get right. Based on these four determinants, we architect various business processes and achieve work tasks. The following are the four determinants:

 Design constructs: Design constructs are based on your firm's competitive and market positioning and strategy. Design constructs emerge from the deployment of capabilities that collectively define a firm's business model and orchestrate how the firm will structure itself.

 Extent and quality of intelligence: The extent and quality of intelligence comes from the core intelligence-centric methodologies. It can be viewed as using the best algorithms for a particular problem set (and the available data; see below), with both effective and efficient training, which results in achieving training goals in a timely manner and with higher precision and recall. Since intelligence manifests at both the artifact level and the networked level connection of artifacts, the extent and quality of intelligence is relevant at both levels.

 Sequence of intelligence and action: The sequence of intelligence and action means that such sequence chains are well defined, optimally placed, stacked for maximum efficiency, and flawlessly integrated. It implies that intelligent software and non-intelligent software (or electric or mechanical systems) function in an integrated, harmonious, synchronized, and efficient manner such that machines accomplish work tasks successfully and optimally.

 Data: Data refers to a firm's ability to have quality data that it can feed to its learning algorithms. Both the quantity and quality of data are important. In addition, the span of data—that means the reality that data covers—is also significant. To clarify the span of data, think about a piece of relevant reality for which you do not have data. Without that data you have no understanding of the reality. As you learn more about the reality, you model the reality with a meaningful representation in terms of variables and features. This multidimensional view of the reality is what the span of data refers to.