Reinventing Projects: A Turquoise AI Project
How the evolution of organizational structures and their learnings can make AI projects not only more successful but also more meaningful.
The colorful evolution of organizations
Starting with red! Cruelty at its best?
Anthropologists say that humanity started to form first organizations around 10 000 years ago. Small conquering, usually crude and violent armies provided an obvious advantage over not organized, weaker neighbors. These military groups were arranged as a wolf pack — with a dominant alpha male, command style, using fear to control underlines.
Frederic Laloux, author of “Reinventing Organizations: A Guide to Creating Organizations Inspired by the Next Stage of Human Consciousness ” calls this type of organization “Red”. Organizations managed with that style still exist in the modern world. Street gangs or mafia are examples.
Amber and the Great Wall of China
Agrarian societies raised organizations to the next level — so-called Amber. The more psychological approach was applied to maintain the strict hierarchy, control, and discipline.
Do what is good for the organization/society and you will be rewarded (maybe not in this life). Play against us and you will be excommunicated.
Amber organizations were better scalable and allowed to deal with much more complex tasks than ever before. Pyramids, Great Wall of China, European cathedrals are just examples. Nowadays Amber organizations can be found in the military or religions.
Orange and the importance of a reward
Renaissance and industrial revolution were another breakthrough moments, which caused that humanity started to consider organizations as machines. Scientific and empirical approaches and effectiveness replaced morality.
The main criteria for decision making in Orange organizations is the highest reward.
Innovation, accountability, meritocracy, budgets, KPIs, targets, etc — these are typical words or ideas which describe an Orange organization. All to achieve the best result in a socially acceptable way. Most of the modern organizations including international concerns are Orange ones. The machine-like Orange approach caused a tremendous development of humanity and unprecedented prosperity in the last 250 years. Unfortunately, this concept is also not perfect. Personal well-being and conditions are often affected as the priority is an organization’s success. Rate race is just an example.
Frederic Laloux in his article ”The Future of Management is Teal” mentions also more massive or even global scale side-effects: corporate greed, over-consumption or reckless exploitation of the planet’s resources and ecosystems. It seems that despite obvious advantages Orange organizations do not provide necessary long term sustainability to humanity.
Green means cooperation and responsibility
Paradigms of equality, solidarity, tolerance, and cooperation are characteristic to Green organizations, which appeared with post-modernity. Most of NGOs but also some universities and commercial entities can be considered as Green ones. Social responsibility is a core of their mission — even if achieved with high cost and efforts. The voice of every member is equal and should be heard.
A metaphor for this type of organization is a family.
Examples are still relatively small but a growing group of organizations seems to evolve above the Green approach. They are called Turquoise or Teal ones.
Three main ideas characterize the next level management style:
· Self-management
· Wholeness
· Evolutionary purpose
One could call them dreamers or radicals but there are more and more managers who set their companies or non-commercial organizations into this concept. It seems that achieving “deep turquoise” is a difficult goal, which requires a mindset of revolution and an organization’s culture reset.
A transition from Orange or Green to Turquoise can be a long term process. The consequence is that it is rather applicable to entities with an expected longer lifetime. There is no time for a long-lasting change if an organization would exist only short.
Well — unless such a short-term organization, for example, a project with its team, is created from scratch as a turquoise one…
A turquoise AI project
I became a member of an AI challenge at Omdena, which was supposed to develop an AI model to prevent fire outburst of forests. The goal has been to detect trees below power lines as they create a risk of fire. The AI model is supposed to analyze satellite images to help with the identification of risk areas. The project is executed by a community of volunteers. Geographically spread all over the world. Driven with no financial motivation and based on collaboration rather than competition. Without a formal hierarchy.
Let us see, is it a turquoise project?
Self-management
Have you ever read one of Stephen Hawkings’ books? How he explains the creation of the universe out of nothing? Matter and symmetric antimatter appeared out of nothing and started to expand just after the Big-Bang around 14 billion years ago. An interesting process was launched just after. The unorganized and chaotic matter at the beginning started to aggregate and form more complex and heavier particles. It took some time but the structured universe including the Earth with leaving creatures and very complex particles like DNA is considered as an outcome of this matter’s self-organization process.
How does this relate to the AI project?
The project team I joined was at the beginning like the chaotic, boiling soup of hadrons at the early age of the universe. A group of people who never met each other, communicating via the Internet, with no formal structure nor a description of responsibilities. A set of unorganized hadrons, which coincidentally appeared in the same “project universe”. The goal of the project was known, a description of the AI model to be developed and its purpose. The communication platform was also provided — an equivalent of basic physics laws which allowed the matter to interact and group as a consequence.
Starting with research
How could we approach the issue? What tools and methodology could be applied? Studying and sharing interesting resources and inspiration. The intensive and still rather unstructured communication was mainly through open messages.
Stephen Hawking was an agnostic and believed that God was not necessary to create the universe. Before joining Omdena, I participated in yet another very similar project, which did not manage to transform from this chaotic phase to another, self-organized one. The difference at Omdena was the presence of “a god’s hand”. This means that the leader or creator of the project made some limited but crucial intervention and set basic rules like weekly common calls to review progress or replacing not active community members and suggesting various roles in a loose manner.
“Take a role you believe fits best to your personal skills and ambitions but please contribute to the community.”
It is also possible to change a role during a project lifetime.
Once concepts and possible approaches were identified so-called “tasks” were defined. Team members assigned themselves to preferred tasks. Sub-communities set individual communication channels, cooperation schemes, methods and smaller work packages, which were further self-assigned by community members according to their individual preferences and capabilities.
The project transferred from the early “After Big-Bang” phase to self-structured development one. Individual communication between team members became more intensive as coordination of development efforts became more important. The system was based on peer relationships, which should not be understood, as in the case of Green organizations, as full equality. Practice shows that efforts to make everyone equal often lead to bog down in debate and factionalism. Leaders and mentors are needed but in a turquoise organization, they are naturally identified and thus automatically accepted by the team. Combination of the self-management and self-organization concept with agile methodology (weekly sprints with regular progress overviews) resulted in surprisingly fast progress and creativity regarding solutions.
Cosmologists say that there might be also other, parallel universes. We may live in a multi-dimensional reality. We just cannot see and experience these. At least without a God’s intervention… Other similar projects but working on different problems were managed by Omdena in parallel. Some of the outcomes or discoveries achieved in a project can be useful also for another one. One project can benefit from the other. Sometimes challenges are similar and coordination of efforts between projects may result in synergies. The “god’s hand” is necessary to establish communication channels and contact team members from parallel “project universes”. The project leader may also need to impose some standards within a project so that blocks developed by tasks fit each other and can finally form the expected complex construction.
Wholeness
Wearing “professional masks” is typical for Orange but also Green organizations, which encourages people to show only their narrow “professional selves”. Wholeness in a turquoise organization is understood as creating and supporting environments wherein people feel free to express themselves. This results in unprecedented levels of energy, passion, and creativity. Orange organizations, including the educational system, which are most common nowadays, create environments which are sometimes difficult to approach but once entered are very difficult to leave. It may be challenging to get a position if you have little experience. It may be difficult to change your profession if you have no required education and some years of experience in a current position. A project which combines a socially useful but not necessary commercially oriented goal with educational mission attracts highly motivated people. The non-financial award which is the possibility of self-development and hands-on experience regarding Machine Learning in a supportive, competition-free environment provides mentors and the possibility to cooperate with other people with a similar mindset is a strong magnet. A magnet that draws people sometimes with little experience regarding AI/ML, sometimes without formal computer sciences education but with the strong feeling that they want to contribute to the community’s goal and develop themselves. This means people who often drop their professional masks and break barriers existing in the orange world.
Evolutionary purpose
Frederic Laloux explains that turquoise organizations base their strategies on what they sense the world is asking from them. Plans, budgets, targets, KPIs and incentives are replaced by agile practices. Paradoxically, by focusing less on the bottom line and shareholder value higher efficiency, also understood as financial performance, is often achieved. Self-definition of a purpose and strategy is less relevant in case of a project, which has a defined goal and relatively short duration. However, it was not defined how to achieve the goal. Methods and tools were not imposed. The community proposed several possible approaches and ideas and selected those to be tested. The possible alternative solutions were developed and tried in parallel — as separate tasks managed and executed by self-assigned community members. The final and best solution was selected as a result of the comparison of tested approaches.
Reinventing AI projects
Commercial companies, including the biggest ones like Google, Facebook, Apple, got really interested in AI/ML solutions since 2012 when some spectacular proves of possibilities of this technology were announced.
Unprecedented involvement of resources available for big commercial companies in the development of AI technology and solutions takes place. Commercial applications of AI solutions become more and more common. Every mobile phone has some. Autonomous cars are not science-fiction now. Scientists are also very active. The software of the Large Hadron Collider in CERN, where scientists try to prove some of Stephen Hawking’s ideas mentioned in this text, was replaced to an AI-based one.
Does it mean that the future of AI is drawn in orange colors? Do only commercial and scientific organization have potential and resources to execute more complex and not trivial AI linked projects? How about socially useful but not necessary commercially spectacular AI applications like fighting hunger, preventing forest fires or fighting PTSD? Maybe turquoise projects driven by communities is the answer…
