Making a better world. One prediction at a time.

Decision Cycle

Human decisions and actions have a “cycle” – the Decision Cycle. It is universal and applies to every decision made or action taken, whether as an individual or as an organisation. What we don’t sometimes realise is that every decision is based on a prediction, and these predictions are founded on data. It is the accuracy and utility of these predictions, and the positive outcomes that they generate, that allow us to survive and then thrive.

Throughout human history our predictions have been made based on Big, Deep data obtained from our senses – what we can see, hear, taste, touch or smell in an uncertain, constantly changing environment. Human Intelligence has evolved to efficiently process and store this “analog” data, look for patterns then use those patterns to make predictions that are the motor for our decisions and actions, which we live or die by. Importantly, Human Intelligence is designed to extract meaning and value from the patterns seen in data.

Unfortunately, we have not evolved to process and store the vast amounts of symbolic, digital data generated by the Data Revolution. Put simply, we don’t “speak” database, and, even if we did, we process digital data so slowly that we could only access a small fraction of the available data. In this new, digital world, it is to electronic storage and processing we look, replacing Human Intelligence with Artificial Intelligence for the creation and implementation of prediction models and, potentially, corresponding automated decisions and actions. Although this intelligence is uniquely equipped to detect patterns in vast amounts of digital data, it does not know what these patterns mean or their potential value.

To overcome the weaknesses of each type of intelligence and take advantage of their strengths, a new kind of intelligence is required – Hybrid Intelligence – that optimally combines and integrates both Human Intelligence and Artificial Intelligence into the construction the Decision Cycles that determine the success of your business.

Human Intelligence
Artificial Intelligence
Hybrid Intelligence

Human
Intelligence

01 Data

Pros: Efficient in processing and storing sensory data (visual, auditory, etc.).

Cons: Very limited in processing speed for symbolic data (e.g. 50 bytes/sec for text). Even more limited in processing digital data.

02 Value

Pros: We can estimate the potential value of data through meaning, according to the decision that needs to be made.

Cons: It is very difficult to estimate the value of large volumes of digital data.

03 Prediction

Pros: With sensory data, we are able to make precise predictions based on scarce data and extrapolate them to new situations with great skill.

Cons: Our decisions are subjective and biased; our brain, which produces mental predictive models, is a black box.

04 Decision

Pros: We use our intuition and experience to balance different objectives.

Cons: Our decisions are subject to multiple cognitive biases; we become confused when there are multiple alternatives; our decision-making can be too slow.

05 Action

Pros: A large number of actions to be performed require human intervention.

Cons: If there are many actions to be performed, the process becomes slow.

06 Result

Pros: Able to define the goals of the decision-making process, the corresponding actions, and KPIs.

Cons: It is difficult to measure the result of the action or to determine if it worked, if the impact is implicit in the digital data.

Artificial
Intelligence

01 Data

Pros: Efficient in processing, storing, and integrating any structured or unstructured digital data that represents texts, images, voice, spatio-temporal data etc.

Cons: The data has to be in digital form.

02 Value

Pros: None.

Cons: It cannot infer the meaning or potential value of the data it is using.

03 Prediction

Pros: Can make predictions based on Big, Deep digital data containing a large numbers of predictors. White box models allow for an analysis of the relative importance of each predictor. Is objective if the data is not biased.

Cons: Does not include intuition or experience. In general, needs large quantities of data. Black box models make interpretation of results difficult.

04 Decision

Pros: Can carry out quick and objective decisions.

Cons: Difficult to incorporate ethical elements; has no intuition and experience; cannot balance different objectives.

05 Action

Pros: Many actions can be automated leading to efficiencies and a better cost/benefit.

Cons: Many actions may require human intervention.

06 Result

Pros: Given a KPI, the relationships between the KPI and the prediction, decision-making, and action that the model took as a result can be tracked and updated.

Cons: It is not capable of creating its own KPIs.

Hybrid
Intelligence

01 Data

Pros: Efficient in processing and storing any type of data, both digital and sensory, and in multiple formats.

Cons: None.

02 Value

Pros: We can estimate the potential value of data through its meaning and the decision that needs to be made.

Cons: None.

03 Prediction

Pros: Can include both digital and analog data, considering large numbers of potential predictors and combining both human intuition and experience with digital data-based AI models.

Cons: None.

04 Decision

Pros: AI models can provide objective evidence for decision making while human intuition and experience can be used to balance different objectives and to judge the cost-benefit of automating decisions.

Cons: None.

05 Action

Pros: Actions can be automated if possible and when necessary

Cons: None.

06 Result

Pros: KPIs can be created, and the relationships between a KPI with the prediction, decision-making and action taken as a result can be tracked and updated.

Cons: None.