The convergence of human and machine decision making: in sync or diametrically opposed?
Imagine you’re going out to buy a new toaster. You are presented with a huge variety of brands, colours, features, and costs. You have the option to engage with a multitude of stores in person, online, phone, PC, online chat, or email, and you need to fulfil your inherent societal and emotional desires. If you are a true toaster connoisseur (but also on a budget), you will want to make the right choice.
In an ever expanding digitised world we are constantly presented with vast amounts of information and complexity when making even the most basic decisions. Perhaps in the past it was easier – you went to the electrical store and chose the cheapest of the three options that the store person recommended, and then you went home and did something important. But being the toaster connoisseur that you are in an ‘everything digital’ world actually makes the decision more difficult.
With easy access to endless amounts information, product data, reviews and opinions, in addition to the ability to access to more stores and channels than ever, making sense of all the available information to make the right decision for your individual circumstance requires not only analysing a large amount of the information (rationalisation), but also relying on your experience, emotions and judgment (intuition).
The discipline of gathering and structuring complex information for the pursuit of knowledge lies at the very heart of the scientific method, and with National Science Week upon us it is a timely opportunity to inspect how decisions can be made to capture and utilise knowledge.
The nature of decision making has changed. In comparison to decisions in our personal lives, consider the similarities but heightened complexities of decision making in business. Critical decisions such as how an organisation decides when to maintain industrial equipment, which product and pricing combination to display to individuals on a website, and deciding within which channel a person would prefer to interact are core points of value creation. These kinds of decisions are made every day, perhaps millions of times per day.
How effective an organisation is at making informed and timely decisions directly affects how an organisation performs.
Decision making in an organisational environment has changed significantly with the evolution of computing technologies. With the adoption of technology we have the ability to:
- generate data and digitised information from a vast array of sources
- consume a large volume of information quickly to inform decisions through the use of distributed computing, cloud computing, big data technologies, and real-time technologies
- make informed, data-driven decisions with the use of statistical analysis and reporting, predictive modelling, and classification
- create self-learning to make decisions with far more accuracy, complexity and flexibility than ever before with advancements in AI and machine learning
- read and interpret human communication and emotions using cognitive technologies.
These concepts have parallels to how humans make decisions. We:
- search for information
- collect, structure and understand information
- evaluate options and make decisions
- evaluate how well the decision performs and learn from the decision
- interact with people to understand, reflect and read emotional responses.
Humans rely on judgment-based decision making that is informed with emotion, experience and reasoning. For example, you might not know one brand of toaster from another, but experience tells you that the well-known brands that offer a larger range are likely to produce a higher quality product. These brands are likely to be more successful due to fewer returns and better customer loyalty, and you reason they will have more experience designing and manufacturing similar products. They might position the product to appeal to your desire for societal acceptance by indicating that it is highest quality product made.
Technology is now able to capture knowledge and experience in the form of trained machine learning models (defined problem sets with accurate prediction), neural networks (which attempt to replicate how human brains learn), cognitive reasoning based on knowledge maps that can abstract information into knowledge so that it can be queried, and in the not too distant future relational reasoning. This allows a whole realm of organisational decisions to be made that are massively scalable and highly accurate while still constantly learning.
In National Science Week there is a unique opportunity for Australian organisations and the general public to recognise and harness the impact of machine generated and fact based decision making these emerging technologies enable. Not only do these improve organisational performance, but also in the broader sense increase our national productivity, improve societal outcomes, and combat the notion of a post-truth world.
And to buy better toasters.