Progress in healthcare delivery and many other domains and industries is contingent on organisations that are capable of absorbing and learning from increasing volumes of data, and capable of integrating the resulting knowledge with the tacit domain expertise and contextual awareness of humans. This article introduces the notion of human scale technology to describe the characteristics that enable tacit knowledge and digitized explicit knowledge to flow between humans and software systems, and it elaborates the role that agent based formal semantic models and non-hierarchical governance structures can play in this context.

Introduction

In the healthcare sector for example, services are coordinated and delivered via medical practitioners, via specialised clinical staff and administrative staff, and via a growing number of supporting software systems. Whilst the level of automation is rising in many domains [1], human tacit knowledge, situational awareness, and the ability to develop trusted relationships amongst peers and with patients are critical elements of optimal service delivery. The overall confidence of patients in the healthcare delivery system is a function of the levels of trust in clinicians and in the systems and tools used by clinicians and patients.

1. Learning organisations

Feedback loops of information flows between agents are the atoms of organisational learning. The SECI (socialisation, externalisation, combination, internalisation) model [2][3] is a useful conceptual tool for extending the concept of continuous improvement into the realm of knowledge intensive organisations.

Concrete SECI knowledge flows can be visualised and formalised with the resources, events, and agents (REA) paradigm [4], leading to representations that are easily understandable by humans and at the same time easily processable by software tools, as illustrated in figure 1.

A national or regional healthcare delivery system is an example of one of the most complex systems operated by humans. Some aspects of such systems are the result of deliberate design, whereas most aspects are the result of cultural evolution under externally imposed constraints. The learning potential of human institutions is defined by the tacit knowledge of the people that are part of the institution, by the understandability and adaptability of the designed aspects of the institution (including policies and technological systems), and by external constraints that are imposed on the institution (in particular access to resources).

1.1. Complexity level 0

One of the simplest possible learning systems is a system of two agents a₁ and a₂ that can process three categories of events e₁, e₂, and e₃ and store information about these events in a suitable information resource structure r:

1. Agent a₁ triggering event e₁, and agent a₂ storing information about the occurrence of e₁ in r, replacing all prior stored information about events
2. Agent a₁ triggering event e₂, and agent a₂ storing information about the occurrence of e₂ in r, replacing all prior stored information about events
3. Agent a₁ triggering event e₃, and agent a₂ responding with r – information about the stored event

The learning challenge at this level of complexity is limited to the error rates of the communication channel between the two agents.

1.2. Complexity level 1

A learning system of some complexity from the perspective of human cognitive limits is a system of two software agents that can process many different categories of events, store structured information about a large number of events, and respond to events in context specific ways [5][6].

1.3. Complexity level 2

A learning system of medium complexity from the perspective of human cognitive limits is a system of a human agent and a software agent that can process many different categories of events, remember both structured and unstructured information about a large number of events, and respond to events in context specific ways [7][8].

1.4. Complexity level 3

An example of a highly complex learning system is a system of two human agents that can process a very broad range of different categories of events, remember both structured and unstructured information about these events, and respond to events in highly context specific ways.

1.5. Complexity level 4

Some of the most complex learning systems involve multiple groups of human agents, and all the interactions between these groups and within these groups. Such learning systems can only be understood by introducing viewpoints, perspectives, and agent motivations as first class modelling concepts [9][10][11].

1.6. Complexity level 5

The most complex learning systems involve multiple groups of human and software agents, including software agents that perform above human cognitive limits, and all the interactions between these groups and within these groups. Such learning systems can only be understood if software agents are able to make their knowledge accessible in human scale representations that respect human cognitive limits [12].

2. Agent based modelling

As highlighted by the SECI cycle, knowledge within an institution accumulates in two places: within the heads of people (tacit), and within knowledge artefacts and software systems (explicit). The MODA + MODE meta paradigm [13][14] is concerned with supporting the SECI cycle with transdisciplinary cultural practices and tools. The core of MODA + MODE consists of two parts:

1. A set of 26 backbone principles (thinking tools) for creating learning organisations and understandable systems that transcend established discipline boundaries.
2. The human lens, which is a metalanguage for describing the semantics of complex system behaviour at all levels of scale.

The categories of the human lens are invariant across cultures, space, and time, and hence they are suitable structural elements of a metalanguage for specifications of formal domain specific languages [5] in a multi-agent context, which are needed to formalise the descriptions of systems at complexity levels 4 and 5.

3. Human scale technologies

For at least two decades now software developers and their employers have neglected the role of understandability for humans. The result is a web of technological dependencies that no one understands and that cannot easily be analysed in terms of potential risks [15][16].

The risks associated with opaque systems are not limited to classical software systems. Artificially intelligent (AI) systems, and the way in which they are currently designed, further grow the web of dependencies, complete with naïve and simplistic assumptions about human nature and economics baked in [17].

Human scale computing [18] can be understood as the elaboration of the role of cognitive characteristics of humans within ergonomics. Human cognitive limits must become a primary concern in the design of human institutions and technologies [12], in much the same way that human scale physical dimensions and characteristics have shaped the discipline of ergonomics.

In an increasingly software and data intensive human world the objective of human scale computing is to improve communication and collaboration:

1. between humans,
2. between humans and software systems,
3. and between software systems.

Our current technologies and communication tools hardly meet any of the human scale computing criteria to a satisfactory level. I believe that the human lens is an appropriate foundation for further work towards human scale computing.

4. Organisational structures

All effective approaches for continuous improvement [19] (such as Kaizen, Toyota Production System, Waigaya, etc.) and innovation (Open Space [20], Manifesto for Agile Software Development, collaborative design, etc.) share one noteworthy common principle:

The belief in the existence and relevance of social hierarchies must be suspended

This observation is backed up by evidence from thousands of organisations that strive to improve or establish a culture of innovation. By definition, hierarchies confer power on specific groups and individuals, with immediate effects on the ability of a group to learn and adapt to a changing environment. Any form of hierarchy or power indicates dampened feedback loops. Power can be understood as the privilege of not needing to learn.

Research of highly competitive Western cultures [21] demonstrates that the social game known as capitalistic economics is a game of luck. Within that game, success has nothing to do with value creation for society and everything to do with social manipulation skills and lack of empathy [22][23][24].

As long as our economic paradigm rewards social gaming, individuals can improve their odds of success by adopting psychopathic behavioural patterns, and by claiming and taking credit for the work of others. Depending on one’s level of empathy, beyond the façade of social success, mental health suffers more or less in the process.

An alternative approach is for a team to agree on non-conventional measures of success, and to work together as a collaborative team to share knowledge, resources, opportunities and success, and by removing all forms of in-group competition and hierarchical structures, to shift the odds for an entire group of people. Given the level of unproductive in-group competition in hierarchical teams [25], non-hierarchical teams have a clear collaborative edge and are well positioned to thrive [26].

The team approach is better for human mental and physical health [27], and it also allows a group to be more selective in terms of where to look for opportunities and how to contribute to society. Problems with hierarchical forms of organisation result from cultural inertia [28] and from the extent to which humans are culturally programmable [29][30].

5. Competency networks

The competency network within the organisation is the union of all the multi-dimensional domain-specific competency rankings that individuals allocate to the other members within the group [31]. It is the only social structure that directly supports the purpose of an organisation.

The existence of competency networks contradicts the simplistic claim that a lack of hierarchy leads to chaos and dysfunction. However, removal of an established hierarchy does not automatically result in a well-oiled competency network. Cultural inertia [28] can keep fear, mistrust, and in-group competition alive, and easily leads to the emergence of new oppressive hierarchical structures.

All healthy and resilient institutions have a well-functioning competency network [26][32]. A good way to understand competency networks is via the notion of trustworthiness and the nurturing and maintenance of trusted relationships [33].

A competency network is the graph of experience-based pair-wise trustworthiness ratings in relation to various domains between the members of a group.

The trustworthiness ratings in a competency network are tied to specific pairs of individuals, and by definition they are not directly transferable and never aggregated into any global ranking. The notion of competency networks is inspired by the correlation between software system structures and the communication patterns between human software developers observed by Mel Conway in 1967 [34].

6. Conclusion

The real opportunity for human society and human organisations lies not in the invention of ever “smarter” forms of in-group competition, but in the recognition of human cognitive limits, and in the recognition of the extreme value that resides in competency networks.

For the first time, the age of digital networks enables us to construct cognitive assistants that help us to nurture and maintain globally distributed human scale competency networks – networks of mutual trust. It is time to tap into this potential and to combine it with the potential of zero-marginal cost [35] global communication and collaboration.

All successful non-hierarchical organisations replace management hierarchies with a simple advice process [26] that establishes the vital feedback loops that enable the organisation to learn and adapt in a timely manner, even in a highly dynamic context.

References

1. Moreno-Conde A, et al. Clinical information modeling processes for semantic interoperability of electronic health records: systematic review and inductive analysis. J Am Med Inform Assoc, Oxford University Press, 2015;22: pp. 925–934.
2. Takeuchi, Nonaka, The New Product Development Game [Internet], Harvard Business Review, https://hbr.org/1986/01/the-new-new-productdevelopment-game, 1986.
3. Nonaka, Toyama, Hirata, Managing Flow: A Process Theory of the Knowledge-Based Firm, Palgrave Macmillan, 2008.
4. Hruby P., Model-Driven Design Using Business Patterns, Springer, 2006.
5. Bettin J. and Clark T., Advanced modelling made simple with the Gmodel metalanguage. Proceedings of the First International Workshop on Model-Driven Interoperability, ACM, 2010; pp. 79-88.
6. Bettin J. and Clark T., Gmodel, a language for modular meta modelling. In Australian Software Engineering Conference, KISS Workshop, 2009.
7. Murray J.H., Inventing the Medium: Principles of Interaction Design as a Cultural Practice, MIT Press, 2011.
8. The potential and limits of clinical decision support systems, CIIC unConference [Internet]. Auckland, Sept 2016: https://ciic.s23m.com/expected-results/ciic-3-september-2016-auckland/.
9. Senge P.M., The Fifth Discipline: The Art & Practice of The Learning Organization, Currency, 1990.
10. Neurodiversity – The Core of Creativity, CIIC unConference [Internet]. Auckland, December 2016: https://ciic.s23m.com/expected-results/ciic-3-december-2016-auckland/.
11. Design and development of tools for effective self-care, CIIC unConference [Internet]. Auckland, March 2016: https://ciic.s23m.com/expected-results/ciic-5-march-2016-auckland/ .
12. The essence of humanity: interaction and collaboration of humans and intelligent machines, CIIC unConference [Internet]. Auckland, Sept 2017: https://ciic.s23m.com/expected-results/ciic-16-september-2017-auckland/.
13. Bettin J., Model Oriented Domain Analysis & Engineering, Thinking Tools for Interdisciplinary Research, Design, and Engineering, CIIC Blog [Internet]. Auckland, August 2017: https://coininco.files.wordpress.com/2017/08/moda-and-mode-lenses-and-principles.pdf.
14. Bettin J., Model Oriented Domain Analysis & Engineering. Domain Engineering Product Lines, Languages, and Conceptual Models. Berlin Heidelberg: Springer Verlag, 2013; pp. 263-290.
15. Foote B., Yoder J., Big Ball of Mud, Fourth Conference on Patterns Languages of Programs, Monticello, Illinois, 1997.
16. Bettin J., Software, Engineering, Artefacts, Language [Internet]. Proceedings of the SEMAT (Software Engineering Method and Theory) Workshop Zurich, March 2010: http://semat.org/proceedings-of-the-semat-zurich-workshop .
17. O’Neil C., Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy, Crown, 2016.
18. Human scale computing, CIIC unConference [Internet]. Auckland, June 2017: https://ciic.s23m.com/expected-results/ciic-3-june-2017-auckland/
19. Deming W. E., Out of the Crisis, Massachusetts Institute of Technology, 1982.
20. Owen H., Open Space Technology: A User’s Guide, Berrett-Koehler Publishers, 2008.
21. Pluchino A., Biondoy A. E., Rapisardaz A., Talent vs Luck: the role of randomness in success and failure, arXiv:1802.07068v2 [physics.soc-ph], February 2018.
22. Babiak P., Hare R., Snakes in suits: When psychopaths go to work, Harper Business, 2006.
23. Long S., The perverse organisation and its deadly sins, Karnac Books , 2008.
24. Long S., Socioanalytic methods – Discovering the hidden in organisations and social systems, Routledge, 2013.
25. Graeber D., Bullshit Jobs : A Theory, Penguin, 2018.
26. Laloux F., Reinventing Organizations, Nelson Parker, 2014.
27. Bowles S., Gintis H., A Cooperative Species, Princeton University Press, 2011.
28. Kegan R., Laskow Lahey L., Immunity to change – How to overcome and unlock the potential in yourself and your organisation, Harvard Business Review Press, 2009.
29. Tomasello M., The Cultural Origins of Human Cognition, Harvard University Press, 1999.
30. Girard P., Pavlov V., Wilson M. C., Belief diffusion in social networks [Internet]. University of Auckland, 2015 : https://www.cs.auckland.ac.nz/~mcw/Research/Outputs/GPW2015.pdf.
31. Bettin J, Designing filtering, trust building, thinking, and learning tools for distributed high-performance teams, Proceedings of the HINZ Conference Rotorua, November 2017.
32. Wilson D. S., Does Altruism Exist?, Yale University Press, 2015.
33. Bettin J, Elliffe M., Improving Interoperability and Trustworthiness of Healthcare Data Repositories. Proceedings of the HINZ Conference Auckland, November 2016.
34. Conway M, Conway’s Law, 1967 : http://melconway.com/Home/Conways_Law.html.
35. Rifkin J. The Zero Marginal Cost Society. St Martin’s Press, 2013.

 

---

Venues
Dates and times