Note: this is Part 2 of 2 introducing The Data Science & AI Hierarchy of Success. To Read part 1, “Fostering a Data-Driven Culture: The Data Science & AI Hierarchy of Success”, click here.
The prequel to this post established mindset and culture, in the form of Data Literacy and Democratization, as the requisite foundation for achievement at the above levels of the Maslow-inspired framework, The Data Science & AI Hierarchy of Success. This post picks up from the Information level and concludes with Data Science and AI.
Level 3: Information
Information sits here physically or virtually with incredible potential for intelligence (the next level) or disaster (when information is unavailable). How the information is stored and if it is within a well-designed architecture determines numerous factors such as whether information is forced to wait and be sent in a batch or can flow in real time to the intelligence level. This layer also has enormous cost and security implications for on-prem and cloud solutions which are beyond the scope of this blog. For more information, research data engineering and data architecture.
Level 4: Intelligence
Intelligence, both business and data intelligence, should not be discounted in the era of Data Science. Intelligence separates correlation from causation, infuses meaningless data points with stories, and provides insights to leaders who make business decisions at all levels. When done correctly, it is a powerful force behind analytics-driven enterprises. However, when done incorrectly, it is also the level at which spurious correlations occur (such as the statistically significant increase in ice cream sales correlating with murder). As with the ice cream-murder real world example, success at this level depends on individuals who have a depth of understanding of how the data was created, its lineage, and how will be used.
Level 5: Data Science & AI
The key difference between this hierarchy and its predecessors at this level is the positioning of Data Science adjacent to AI as opposed to below it. Why? Many AI applications including RPA (robotic process automation) can be built directly on top of Business or Data Intelligence. Data Science and AI also share methodologies and applications which can be considered as belonging to either or both domains. AI, or more specifically, Artificial Narrow Intelligence (think FaceApp or AIPortraits) as opposed to Artificial General Intelligence (i.e. Skynet and HAL), has incredible potential within the enterprise from improving customer experience via conversational AI to creating new product and service offerings (both exemplified by Amazon’s Alexa).
Data Science, or the application of the scientific method to data via computer science, statistics, and domain experience, has the potential to uncover new insights and create opportunities beyond the scope of Business Intelligence. The models it employs allow us to explore the “what ifs” and hypothesize, moving us beyond describing the past to creating informed predictions about the future which take into account complex variables including randomness itself.
Data Science and AI can (and should) be pursued simultaneously based on the foundation outlined in The Data Science and AI Hierarchy of Success. More so, within organizations which have a Data Literate workforce, opportunities to leverage these methodologies for digital transformation and innovation can come from anyone. Anytime. In any part of the organization.
What do you think? Does the Data Science and AI Hierarchy of Success resonate with your experience?