A series of conferences about climate change were organized few weeks ago by the Perelman centre from the Université Libre de Bruxelles. Here are two observations from François Gemenne:

• Climate refugees are most often treated as if they were victims unable to adapt to the climate changes occurring in their region. They play the role of “canaries in a coal mine”, signalling the rest of the World what is to come. We should instead recognize that emigrating is a legitimate adapting strategy to climate change, and that climate refugees are far from being victims without any capacity to act. Displaced populations are only one aspect of the problem. We need to better understand the conditions that lead to the decision to either stay or leave an affected region.
• More a country is polluting, more it has weight in climate negotiations. Model pupils are not the ones who need to change their behaviour, polluting countries impose their conditions for change.

ECCS'10 was held in Lisbon, I saw this huge graffiti by the artists Os Gemeos one evening when I went back to my hotel, also about ethics...

Let me conclude my series of posts on the ECCS 2010 with few notes on power and ethics. You will be convinced by reading my previous posts that complex systems theory has many applications. The discipline is relatively new and we are only at the beginning of discovering the impact it will have on our daily life. As for any science, it can lead either to a better world or a nightmarish society, depending on how we use it. I could feel it throughout the week of the conference. Most scientists advocate the positive impacts of their research, but do see the possible misappropriations. Complex systems theory is central in the development of a more sustainable society, as illustrated in innovations such as smart grids and intelligent transport infrastructure, which happen to be decentralized systems. Complex systems theory provide us with a better understanding of our cultures and the way we operate together, which could help us address societal challenges. It could help us better classify knowledge, plan our cities, simplify our laws, reach altogether informed decisions and preserve cultural diversity. But it could also lead to control over public’s opinion, to a wider social gap between influencers and their followers, to a resilience of monopolistic systems, and to partitioned societies. The choice is ultimately to citizens, and it worries me to see that the vast majority doesn’t care much about recent advances in complex systems theory, as in any science, too busy to cope with what has been already decided for them long time ago in other fields. What is necessary to make that change? In the meantime, it seems to me that scientists are left on their own, their only interlocutors being large corporations and short-sighted politics (who cannot carry any public will on the subject because the public doesn’t care). Scientists improvise the best they can social and philosophical criticism of their own research, but it would much more beneficial to have other disciplines involved.

After the inspiring presentation  by Jeffrey Johnson, “Policy and Design of Complex Systems”, I wondered if a form of ethics in social systems analysis might prescribe the modelling of every person involved , including the creators of the model and their commissioners. Self-interest and personal motivations could then become much more apparent than in current researches.

“There is a power struggle for ‘the Truth’ [...] Scientists must model themselves and the politicians, both are inside the system being studied. Modelling the system can change the system!”

http://vimeo.com/15899828

Applications of complex systems theory in policy making

“Intervention and Policy Making in Complex Socio-Economic and Technical Systems”

by Peter Allen

Smart grids

“Modelling the growth of distributed energy generation for a low carbon economy: Part I the approach”
by Liz Varga

The greatest challenge for the UK government is to ensure that energy demand is met and sustainable over the long-term. This assurance is currently achieved through centralized control over various gas and electricity licensees, including generators, inter-connectors, distributers, etc. The future however is dependent in a significant way upon households and organizations, taking responsibility for their use and, where possible, their generation of energy. This new decentralized model which has the opportunity for self-organization and growth, in the same way as the internet had with the deregulation of telecommunications, is known as the ‘smart grid’ in the USA. The purpose of modelling the power industry is three-fold. First, the power industry cannot be experimented upon because of its scale and the reliance placed upon it by the economy. Second, effects of potential governmental interventions will depend upon the evolving environment and other dynamics existing within the system and so are difficult to predict because of the dependence upon the context of the interventions. Third, is to demonstrate with reasonable probability given the assumptions made and given the effects of noise in the system that desired outcomes relating to low carbon, security and sustainability are achievable. Human or social agents and artificial or smart agents will be modelled at similar levels of abstraction and simulated together, allowing an understanding of the interactions between technology, information, individuals, communities and organizations, building on the use of interpretive agents who learn.

http://vimeo.com/15623991

UK Transport Emission Reduction

Most often in environmental issues, not one but several policy measures need to be implemented in order to get positive effects, and the interaction between these measures becomes key in their success. Finding the set of measures that will lead to the best synergies is a very complex problem indeed. Araz Taeihagh argues that policy makers should let science help them, and use policy models in their design process. His model is based on five types of relation between policy measures: a measure can be required for the application of another (precondition); a measure can facilitate the application of another (facilitation); two measures can work well together (synergy): two measures can produce conflicting outcomes under certain conditions (potential contradiction), two measures definitely produce conflicting outcomes (contradiction).

“A Case Study in the Application of an Agent-Based Approach in the Formulation of Policies for UK Transport Emission Reduction”
by Araz Taeihagh and René Bañares-Alcántara

http://vimeo.com/15624248

Hospital refurbishment

By The Open University

Hospitals are complex: they involve strict codes of operations, nurses, doctors, specialized equipment, operating rooms, and patients of course. They are optimized for maximum effectiveness in healing patients. A single change in their settings can lead to chain reactions and unpredictable consequences in daily operations. How to integrate then planned refurbishments and new requirements such as reduction of energy consumption? The new study of the Open University will attempt to model hospitals using complex systems theory in order to address the challenge.

Complex systems theory applied to hospitals

This is what Alexy Khrabrov and George Cybenko attempted to answer at the ECCS 2010 conference in Lisbon. Here is an introduction that I have slightly edited to make it more accessible.

“Mind Economy: Modeling Influence in Communication Networks with Social Capital.”

by Alexy Khrabrov and George Cybenko

Social scientists, businesses, and governments are interested in summarizing the ongoing social network activity to identify the most influential players capable of creating and maintaining high-impact group behaviours. We would like to have metrics of influence in dynamic systems, and generative models which can explain how this influence is accumulated and maintained. Having identified the “stars” or high-influence individuals, we look at the ways they achieve and maintain their influence, comparing their tweeting behaviour to social capital exchange in proportion to the fans ‘contributions. We propose a family of generative models where social capital is exchanged and generated during interactions, reflecting the players‘utilities – such as self-centred or maximizing group benefit. Using our social capital model, we let a system evolve to accumulate most of the capital in those nodes which can be considered influential. We compare those capital-rich nodes with other metrics of influence and show that our model confirms and explains influence of many important types of players, and reveals the behaviours leading to sustained influence. We model recently uncovered Twitter phenomena such as Justin Bieber‘s ecosystem and other high-intensity processes, showing how efficient star behaviour and group preferences lead to various mind economies in social networks.

http://cs.wellesley.edu/~cs315/Papers/khrabrov-twitter-dynamics.pdf

So, even the seemingly chaotic activities of a teenager can be modelled using economic models in which influence and social capital is the currency. Trends on social networks such as twitter and facebook still seem rather unpredictable. But the interests at play in a mind economy are too big to let it go that way. Complex systems theory will undoubtedly be used in an attempt to better target influencers, and shift our attention to a particular subject. The same theory might on the other hand help us making sure that social networks maintain a certain level of social fairness in the mind economy, and diversity in mindshare.

Simulating opinion dynamics in heterogeneous communication systems

By Walter Quattrociocchi, Rosaria Conte, and Elena Lodi

In this video, Rosaria Conte describes opinion dynamics by means of multi-agent based simulations. Agents (i.e. people) are exposed to different sources of information varying both the contents and the perceived reliability of the messages spread. Agents’ internal opinion is updated either by accessing one of the information sources, namely media and experts, or by exchanging information with one another. They are also endowed with cognitive mechanisms to accept, reject or partially consider the acquired information. The study evaluates the impact that reliable sources and peer-to-peer communication can have on the quality of information.

Simulating Opinion Dynamics in Heterogeneous Communication Systems from Assystcomplexity on Vimeo.

At the third minute of the video, Rosaria Conte starts attacking the Italian media manipulated by Berlusconi. Behind the rather tedious title, this presentation is a fantastic example of political engagement through science, which doesn’t undermine at all the scientific value of the research.

Opinion dynamics

By Anthony Woolcock

Many societies exhibit cultural fragmentation. This is despite individuals trying to reach agreement with those they meet. In the model of Axelrod, individuals that are more similar are more likely to interact (homophily). The mechanism where people become more similar after interaction is termed social influence. Axelrod’s model is interesting because for different parameter choices the opinions of all the individuals will either become all the same (consensus), or frozen fragmented state. A phase transition is observed between these two types of frozen state.

Complex systems theory: opinion dynamics

Bounded confidence model: addressed information maintain diversity of opinion

By Krzysztof Malarz and Krzysztof Kulakowski

Models have already been developed to optimize the frequency of let’s say political advertising on tv in order to get the maximum effect on a public’s opinion (see the Zaller mass opinion model). But they didn’t take into account interpersonal communication, which becomes more and more decisive with the advance of online social media. Krzysztof Malarz and Krzysztof Kulakowski developed a model taking this into account. Their model suggests that individually addressed messages maintain diversity of opinion. See the full paper here.

Model for diversity of opinion

I don’t know many fields that are not confronted with complex systems. Chemistry, physics, maths and all the scientific disciplines I studied at university are dealing with structures that require more than simply understanding their parts. I had the occasion to deal with other types of systems when I was working in business innovation: organisations are made of networks of people, networks of processes and are intricately connected with their economical ecosystem. I have seen many complex systems in culture too: semiotics, linguistics and urbanism just to give few examples. Finally, in the researches I’m currently making on consciousness, systems play also a major part, such as in neuroscience. So, complex systems are everywhere. We are only beginning to understand their mechanisms, and how interdisciplinary their study is. I will publish a series of posts on the subject in the coming weeks, which I hope will interest you as much as I was by the conference.

• Contrary to the widespread belief, brainstorms are more about giving sense to old ideas than discovering new ones.
• Every model is based on assumptions. Paradigms and stereotypes are more abstract models that don’t necessary come with explicit assumptions, which narrows down creativity.
• Creativity is about changing perception. Innovation is about changing things.
• Techniques for workshops:
  • Choose a random organisation having nothing to do with your industry. What is the probability it takes over your business in 10 years? Explain SWOT.
  • Speak about your job without using the most commonly used words to describe it.
  • A paradox is when reality doesn’t follow the model. A new model must then be applied. An ambiguity is when reality can be represented by several models; it is far more common. Creativity is about building new conceptual ‘boxes’ inspired by ambiguities.

Luc de Brabandère is ‘corporate philosopher’ and director at BCG, check out his books here.

• An economy of contribution is an economy where actors contribute without having necessary a measurable counterpart (e.g. money, product). This model is radically different from an economy of exchange and must be taken into account when the externalities underlying an industry become central to its process. In this context, the function of companies is to capture positive externalities.
• An organisation is at the cornerstone of three axes:
  • Patrimony: the former ‘inner’ single reference for a company: legitimacy, expertise, employees…
  • Employability: capacity to inscribe its products and usages in shared values. Employability is related to economical theories of ecosystems in which an organisation is seen as a part of a system, either parasitic or in symbiosis.
  • Interactivity: capacity to adapt to the needs of individuals and communities who will in return support the activities of the organisation. With the rise of social media, brands want to be seen more and more as Customer to Customer ‘CtoC’, like if they were initiated by customers and driven by them.

Mozart: a talent detected in his early childhood

I attended few weeks ago a seminar organised by the Philosophy and Management association in Brussels. It was all about talent, and how the way artists work and manage their career can be a source of inspiration for talent management within business organisations. Pierre-Michel Menger, French philosopher and research director for the CNRS, presented some of his researches in the sociology of work and art. The expertise of Pierre-Michel Menger in both fields led him to very interesting observations.

He first discerned two types of work:

• the ‘labour’: an effort, constraint with a predictable outcome
• the work as a discovery of yourself, the masterpiece of your life. Success in this type of work is more a derivative, not a predefined goal. It is unpredictable.

The later work is influenced by philosophies from the 19th century emphasising the infinite depth of consciousness and the infinite possibilities opened to us. Because of it’s unpredictability, it is a type of work that involves a lot of risks. The prestige and satisfaction you get from the realisation of a masterpiece is immense, but the risk to fail is in equal proportions. Artists are facing this risk in a hyper competitive environment. Differences in revenues between artists is huge, the small number of successful artists take most of the resources, leaving a small portion of revenues to the vast majority. If you look at just the financial situation of an artist, choosing such a career might look like a bad evaluation of the risks. But it doesn’t take into account the non-monetary value of a potential huge gratification, a relative autonomy and the diversity of the tasks involved in the job.

What influences the likelihood to become a successful artist (at least in the narrow sense of social recognition and monetary compensation) is not clear. There is something about artists that cannot be measured, cannot be put in an equation. It is not enough to make studies, work hard and accumulate experiences. You need to have ‘talent’. The same applies to business. Sure, you can find people able to perform a task by looking at their past experiences and by using rational criteria. But how can you detect the collaborators who will go far beyond, surprise you and develop considerably within the company. How can you make sure to invest more in these people and less in the others? McKinsey invented a marketing term to describe the 10% of your employees who will bring the most to the company: ‘talents’. Talent management is controversial and relates to the many meanings of the word ‘talent’.  Pierre-Michel Menger proposes to define talents as people you cannot isolate using predefined criteria or reading their CV. It helps to stick to this definition and to not take into account all the other meanings and judgements that the word ‘talent’ implies. He argues that the only way to detect talents is to compare them between each others. This is why competitions and awards in the art world are so frequent. The jury don’t know themselves what they are looking for, and the outcome is unpredictable.  It is after comparing the contestants that they can see who has a little something more, a higher potential. This is also why I think you start to see more and more contests and game-like workshops in business (hack day for developers, unconferences, Dragon’s Den kind of internal events), to detect talents you could not screen using a formal HR equation.

I think that this approach raises many ethical dilemma.

• On one hand, I want people to judge me on rational criteria that I can understand and act on. It is a system that protects all of us against arbitrary decisions and favouritism. On the other hand, I also want to be judged for who I am and for my talents, independently of any predefined list of criteria.
• Whatever your boss says, it is reassuring to know that he doesn’t judge you but only your work and your performances. With the concept of ‘talent’, suddenly your boss wants to know what you do in your spare time and wants you to reveal yourself, so that he can detect the ‘talent’. It is a much more personal relation. It surely benefits some people, but not necessary everyone. The opposite leads to the same problem but the other way around. You might have a quality that you know could help your career, but an employer looking only at short term figures and performances might not realise it.
• Is it right to invest more in the 10% of ‘talents’ in your company instead of using that money to raise the general level of expertise of the team? To use an example from Pierre-Michel Menger, if a talented researcher wins an award for a paper he wrote, his reputation will get a boost which will convert probably to a higher salary. But the paper he wrote is most likely based on data that ‘average’ workers collected. Is is fair? On the other hand, if there is no incentive for researchers to excel and be noticed, people will stagnate and become demotivated.
• Even small differences between people of the same level in a specific field, like music composition, can generate disproportionate inequalities. If more is invested to a young ‘talent’ who is a little better than the other children of his class, he will quickly gain more experiences and have more chances to explore his talent. He will then have a reputation, which will encourage people who don’t have time or the knowledge in music composition to hire him instead of someone else, which will give him even more experience, and so on. The ‘talented’ person might truly be exceptional, but was it because of his initial tiny competitive advantage or because of the investment from the community? Does it make a difference? Maybe humans are like bees and need an arbitrary hierarchy for their society to work. (Simply accepting this image of the bees is way too sympathetic with established power though.)
• As it has been pointed during the workshop, talent belongs to the category of work that is unpredictable and can be asserted for sure only a posteriori. Is it legitimate to try detecting talents beforehand? This argument is interesting but theoretical.  Of course, people will always try to detect talents, it happens since the beginning of humanity. And even if it is an inexact science, it probably lead to better results for the community than not trying to support its future ‘talents’. Even if it is not always fair for people.

How to resolve the dilemma depends on your vision of society. If you believe that there is a real opportunity for people from all backgrounds to display their talents, then selecting talents by comparing people between each other on non measurable criteria is legitimate. If on the other hand, you perceive the world as being a constant exploitation of the masses by few people in power, every privilege not based on measurable merit is a potential discrimination. Both extremes are false, the world needs both talented artists and hard workers. I personally believe that the key to resolve the dilemma is to offer multiple ways to succeed, in many different ways, with the help of many different groups of people. Diversity lowers the probability of generating systematic discriminations and enables many understandings of what talent means.

If you are preparing an award or competition in your organisation, you need to understand why you the feel the need to do so. If it is at least partially to detect talents, I hope that the points above will help you design the process in accordance to your values and goals. Don’t simply replicate what has already been done, enable participants to show their talents from a difference angle, the winners might not be the ones you expected.

I collaborated with him to prepare the photographs and Polaroids, you can find the details here.

I hope it conveys the message and the exciting mission of the Philosophy and Management association.