Marc Vollenweider is the CEO and Co-founder of Evalueserve.
Artificial intelligence and smart tools are becoming increasingly prominent. Although these tools could be crucial in some cases, real value can only be created when they are combined with the power of the human mind.
Before I set out to convince you about this bold claim, it is important to define some of the basic principles of what I call our “Pyramid of Use Cases.” This four-level pyramid gives us a better understanding of the evolution of big data into coherent information and sound insight, thereby leading to usable knowledge. For øjeblikket, the words data, information, insight, and knowledge are either defined loosely or used interchangeably, leading to significant confusion.
Level 1 – Data: This is raw or cleansed data, basically a collection of facts that does not reveal an overarching meaning without further analysis. In itself, data has no direct value for a decision maker and often requires artificial intelligence and smart tools that can prepare it for further analysis. An example of data is sequences of measurements received from sensors in packaging machines.
Level 2 – Information: This is data that is processed to an extent and put into context, but is still not in the format needed for making critical business decisions. An example of information is the discovery that the data sent by the sensor has five unexpected outliers, where vibrations are stronger than allowed under technical specifications.
Level 3 – Insight: This is what a decision maker is looking for. Insights provide decision makers with the necessary “so-what” to make informed value-adding decisions. Following the previous example, an actionable insight could be a finding that the packaging machine’s vibrations are stronger when it is run with a load lower than a certain weight.
Level 4 – Knowledge: This is the essence of what analytics, and indeed research, should aim for. Knowledge comes from insights that have been made re-usable over time, allowing successful business decisions. An example of a sound decision could be that the packaging machine’s guidance is updated to increase the minimum load weight.
As we move from data toward knowledge, volume reduces, with millions of individual records (data) boiling down to a simple line of advice or guidance (viden). Decision makers want compressed insights and knowledge, not a massive mountain of raw data.
To best depict the evolution from data to knowledge, we have codified the eight steps of the analytics value chain. While the effort and cost is spent mostly at the beginning of the cycle, value is created largely at the end, det er, when a decision is made. I call these steps the Ring of Knowledge.
Coming back to my assertion that real value can only be achieved with a combination of mind+machine, let’s run through the above steps and consider where the mind or machine needs to be applied.
- Gather, cleanse, and structure data (steps 1–2): Dag, the collection and cleaning of raw data relies heavily on machines. APIs can collect and process large quantities of data faster than even the most capable data scientist. Endvidere, intelligent automation makes no mistakes (if programmed correctly) and the data being fed into step 3 becomes vastly more reliable.
- Create level 2 information (step 3): For nylig, machines have begun to play a bigger part in the creation of information. Our investment banking team uses VBA macros that can filter and arrange industry data by market share, country, and product category.
- Create level 3 insight (step 4): Creating insight is about isolating the so-what from the information. This requires the perceptive power and lateral thinking capacity of the human brain. Even the cleverest artificial intelligence application cannot assist with this step yet.
- Deliver insight and make decision (steps 5–6): The value of data analytics is in delivering the right insight to the right decision maker at the right time. Smart tools, such as mobile apps and push notifications, can assist with this process. But when it comes to the resulting decision, human experience and instinct are essential.
- Create and share knowledge (steps 7–8): Endelig, when it comes to creating knowledge, the human mind is needed to crystalize long-term learnings. Dog, storing and distributing this knowledge can be significantly supported by a sophisticated knowledge management platform.
Power of mind machine
I believe the above steps make it clear that mind and machine have to work in tandem to complete the whole Ring of Knowledge and get the most out of data analytics use cases.
Machines, which are now essential for steps 1 og 2 (at level 1 – data), are increasingly becoming more efficient. A level of automation has also been achieved for step 3 (level 2), with machines drawing information from data.
Dog, the human mind is essential for steps 4 til 7, because even the cleverest artificial intelligence applications cannot achieve desired outcomes. Machines can process large volumes of data far more efficiently and effectively than humans can. But the ability to spot and react to new patterns, understand the broader business context, differentiate between correlations and root-cause relationships, and tell machines what to do will fundamentally continue to remain human tasks, at least in the foreseeable future.
Using the Ring of Knowledge will allow you to identify when and where you should seek to automate og when you need to make sure you have the right minds available.
To come back to my original point, decision makers want compressed insights and knowledge, not raw data or even information. Efficiently delivering this is possible only with a balanced combination of mind+machine.
If you want to learn more about how to optimize and implement data analytics use cases, you can take a look at my book, Mind Machine – A Decision Model for Optimizing and Implementing Analytics.