How do we become an AI-driven enterprise? (Part 1)
By Tim Beswick
We are often asked whether becoming an AI-driven enterprise requires something different than becoming a Data-driven enterprise. In this series of blog posts Anderson MacGyver shares her point of view on this topic.
Data-driven enterprise
It is good to start with looking at what a Data-driven or an AI-driven enterprise is. Anderson MacGyver takes a business value centric view and defines a Data-driven enterprise as an organization that unlocks the business value of data in any of the following three ways:
- Enabling digital systems with the exchange of data: Today’s society relies on systems exchanging data. The importance of the exchange of data between digital systems has superseded the importance of spoken and written words between humans. Without the timely exchange of the right data between digital systems, an organization and our society come to a standstill. Data-driven enterprises ensure that the digital systems in their organization and eco-system have the right data, of the right quality, at the right time.
- Deriving commercial value from data: Data on itself can have commercial value. Data-driven enterprises derive monetary value from their data by trading data either for a direct monetary reward or in exchange for other goods or services.
- Creating insights through analysis of data: Data analysis provides insights that can be used for improving customer intimacy, strengthening business control, driving operational efficiency and for the development or as part of new products & services. Data-driven enterprises unlock value from data by using data-driven insights to optimize their business processes and offerings.
AI-driven enterprise
So how does an AI-driven enterprise differ from a Data-driven enterprise? Let us first look at what an Artificial Intelligence is. Anderson MacGyver uses the following terms to define an AI. An AI is a digital system that:
- Has the ability to learn and adapt.
- Can generate output in the form of new data and content (video, image, text, sound, code).
- Triggers or autonomously take actions.
What does this mean in relation to the three types of value from data?
- Exchange of data between digital systems: An AI can be used to optimize the processes that drive the timely availability of data for digital systems. This does not result in new types of value in this domain.
- Commercial value: An AI can be used to generate new data and content that has commercial value. AI becomes the core production engine of your commercial goods that contain data, video, image, text, sound or code.
- Insights: AI can provide dynamic data-driven insights and if desired autonomously act. By autonomously acting an AI often improves process efficiency. The AI generated insights are used for improving customer intimacy, strengthening business control, driving operational efficiency and for the development or as part of new products & services. In these kinds of applications, AI is an extension of using data driven insights to optimize your business processes and offerings.
An analogy to a smart car can help to make the point on dynamic insights and ability to act autonomously more tangible:
- A route planner that calculates the best route to your destination based on a static map and ignoring current traffic situation is not an AI.
- A route planner that dynamically recommends the best route based on the current and predicted traffic situation is not an AI if the recommendations are based on predefined rules (Note: rules can be defined using historic data-driven insights).
- A route planner that dynamically recommends the best route based on the current and predicted traffic situation is an AI if the recommendations are based on predicted future traffic patterns, the system continually learns from historical data and optimizes the route based on multiple, complex factors.
- A system that controls your autonomous vehicle taking the most advantageous route into account, must be an AI if you would want to safely make it past even the first junction. It would have to respond to the ever-changing circumstances on your route.
Autonomously acting is not an absolute phenomenon. It comes in many forms. From a simple trigger in the form of a recommendation to operating and controlling systems without any human intervention.
Summarizing the above, an AI-driven enterprise is an organization that leverages Artificial Intelligence to unlock business value by using digital systems that, based on data, learn and adapt and:
- Generate new video, image, text, sound and code.
- Trigger actions or autonomously act.
Data analytics is typically used as part of an AI system. This implies that an AI-driven enterprise is an extension of a Data-driven enterprise.
Data-to-AI-to-Value
Becoming a Data-driven or and AI-driven enterprise is a journey. Anderson MacGyver uses the terms Data-to-Value and Data-to-AI-to-Value for an organization’s journey to become a Data-driven or an AI-driven enterprise.
In the next blog post we share the good practices that we learnt to apply in these journeys. So, stay tuned!
Interested in further insights into this topic? Join our CIO Masterclass on becoming a scalable, AI driven enterprise on the 13th of November.