Collaboration of human and machine for knowledge work
To create business value from data-driven decision-making, organisations must apply the insight from both knowledge workers and intelligent machines.
The world is seeing revolutionary advances in science and technology, and with the evolution of digital technologies, there is a growing recognition that most workplaces are experiencing change.
The predominant tendency of this change is toward understanding and managing greater complexity, known from a collective perspective as knowledge work.
Consequently, the evolution of digital technologies also changed the landscape and nature of knowledge work, as the growing use of digital technologies has created entirely new business models and ways to create value.
Some of these include control and monitoring through computer-based algorithms (eg, cyber-physical systems such as autonomous vehicles), the proliferation of a connected world (eg, Internet of things creating enabling smart cities), on-demand availability of computing power and data storage (eg, cloud computing) and cognitive computing (eg, artificial intelligence).
Maintaining a knowledge work emphasis in this context is important, seeing as intelligent machines are altering the knowledge creation and knowledge sharing methods in organisations. Furthermore, a key contributor to the viability of intelligent machines (artificial intelligence technologies, machine learning models), is the availability of data that may be applied in computer learning processes.
In addition, big data structures – structured and unstructured − are used to extract value from, and consequentially, data-driven organisations base their decision-making evidence on data, rather than intuition.
Most organisations have more data than they know how to exploit effectively, and to a large extent, there is still gap between impact and big data. Thomas H Davenport and Randy Bean reported that, according to a survey conducted among senior executives from 57 large corporations, organisations are slow to make the shift to a data-driven culture, as 99% of respondents aimed to achieve it, yet only one-third succeeded.
A key opportunity to achieving this impact is to create valuable knowledge from big data, and consequently enabling winning strategies in a competitive world.
This becomes apparent in the different and new ways of engaging customers and harmonising activities, as well as in the fact that some of the world’s largest enterprises conduct business based on the new technologies, challenging conventional divisions of labour between man and machine, consisting of inter-related initiatives such as automation to diminish repetitive jobs, digitisation of work to improve worker efficiency and artificial intelligence to provide more reliable, useful and productive professional work.
Most organisations have more data than they know how to exploit effectively, and to a large extent, there is still gap between impact and big data.
Although the application of artificial intelligence and machine learning proliferate perceptions that humans are obsolete to some extent, the machine learning delivery pipeline still requires human interaction (eg, parameter tuning, training of models or feature engineering).
Additionally, when models need to learn from human preferences such as recommender systems, when security concerns require interpretability of the learning process and outputs, or adapt to users, human input is fundamental.
Therefore, recent research concentrated on interactive forms of learning and machine teaching as it actively utilises human input and in this way, learns from human intelligence (robotic process automation). This enables machines to learn tasks they cannot yet achieve alone, adapt to environmental dynamics, and deal with unknown situations.
In light of deep technological changes brought about in every organisational facet by the emergence of intelligent machines and big data structures, it becomes highly relevant to revisit assumptions about the nature of knowledge work. Researchers found that, to create valuable knowledge from big data, at least four enablers must be considered.
- Data analytics is the process of exploring datasets in order to draw inferences from the information contained in the datasets. What are the regular type of business questions which are imposed on the data? What are the business rules and algorithms that are used by analytical models such as predictive, descriptive, classification logic, etc? What are the specific use cases; ie, the definition of specific situations and business rules in which a product or service could potentially be used?
- Data management as an enabler is the practice of collecting, keeping and using data securely, efficiently and cost-effectively. What are the data processes and the steps used to ensure the management of data as an information asset? What constitutes clear needs definition in order to manage the organisation's information demand and supply? What data model and standard do you use that structures information in such a way that it is easily accessible and well understood to support data-driven decision-making?
- The enabler data platform denotes an integrated technology solution that enables data management for strategic business purposes. Do you have the big data technical capability to store, organise, manage and secure the organisation's data? Are you managing data identification and application in support of the exploitation of data through reporting, analytics and extracts to realise benefit such as monetisation? Do you have sufficient data ingestion capability for the assimilation of application source data within the big data storage?
- Data-driven organisational ethos is an organisation’s commitment to gather data that can foster conclusive decision-making concerning all aspects of the business, ensuring it becomes part of the organisation’s competitive advantage. What tactics does the company employ to drive decision-making from data? Does the organisation foster values, expectations and practices that guide and inform its actions as it pertains to applying data to inform decisions? Has it shared a documented and published data strategy, including a set of goals and a time-boxed roadmap to support an insights-driven organisation?
From an organisational perspective, decision-makers are now empowered to derive actionable insight based on the analysis of big data datasets through advanced analytics. From a knowledge worker perspective, actionable insight and data-driven decision-making are associated with transforming data to knowledge and ultimately to wisdom requiring organisations to better understand fundamental constructs in this context.
Associate professor, Department of Informatics, University of Pretoria
Dr Hanlie Smuts is an associate professor in the Department of Informatics at the University of Pretoria since 2017. During her tenure in industry, her role aimed to deliver consistent customer relevance across all digital touch points, empower customers through convenient and effective self-service, and drive growth through personalised digital offerings. Through a deeper understanding of the digital and adjacent ecosystems, she championed transformation to digital and the need for collaboration in this context. She currently focuses on research in IT and the organisation, with particular emphasis on digital transformation, disruptive technologies, big data management, enterprise architecture and knowledge management. Dr Smuts has published several papers and book chapters in her field of study.