Wednesday, September 22, 2010

The Future of Economics

The recent failure of classical economics to predict and manage the catastrophic failure of the world’s financial system has triggered a re-evaluation of the whole basis of current economic theory, which has been applied to sustain capitalism for the last 100 years.

By the end of the 20th century traditional economics was dominated by the classical paradigm based on notions of rational consumers making rational choices in a simple supply/demand world of finite resources, with prices constrained by decreasing returns; all driving the economy to an optimal equilibrium point.

Twentieth century economists had finally realised their dream of creating a rational, rigorous and well-defined mathematical model for describing the workings of the global economy. This standard model has been applied by business leaders, finance ministers, central bankers and presidential advisers ever since.

Up until recently classical economic theory has appeared to work adequately by a process of trial and error. In times of growth people are generally optimistic and the theory describes reality reasonably well. But in extreme circumstances panic quickly spreads and the theory fails spectacularly, amplified by the performance of the quantitative risk algorithms beloved by hi-tech stock market traders.

Unfortunately such a clockwork model has proved over the last four decades to be seriously out of synch with reality, as global markets have been roiled by a series of disastrous credit, market, liquidity and commodity crises. The predictions of the standard model have failed to match real world outcomes, generated in succession by the Savings and Loan, Asian, Mexican, Dotcom and now GFC bubble disasters.

In this latest incarnation of excess greed debacles, high risk mortgage loans were repackaged many times over into opaque risk financial instruments, such as Collateralised Debt Obligations or CDOs, which ended up through an unregulated banking system in the portfolios of nearly every bank and financial institution around the world. Because of lack of controls, members of the shadow system such as hedge funds and merchant banks borrowed scores of times their own worth in cash. When the CDOs finally failed, the losses rippled through the world economy. The banks stopped lending, leading to further business failures and investors were then forced to sell previously sound stocks causing a stock market crash.

But this crash was far more serious- perhaps even more so than the Great Depression, as it could not be contained within borders as easily or so simply solved by pump priming mass lending and job creation programs. Now we’ve seen the biggest banks, car manufacturers, miners, energy suppliers and national economies toppling like dominoes around the world, under trillions of dollars of debt.

The current global interventions have now staunched the haemorrhaging but not cured the disease.

The stronger economies of China and south east Asia, Brazil and Germany, less affected by the carnage, have bounced back. But the European economy is still fragile, with Greece, Spain and Portugal and other smaller nations struggling to contain debt; while the recent G20 summit in Toronto failed to enforce the rigorous regulation and improved economic governance previously mandated. The US recovery is also weak, with the latest OECD report predicting that the US employment rate will not fall to pre-recession levels before 2013.

In fact a number of interdisciplinary thinkers, starting in the seventies, began to question the credibility of the entire basis of the classical economic model, likening it to a gigantic academic think tank experiment rather than a serious science. And it gradually began to dawn on this group that at a number of the key premises or axioms underpinning the existing model were seriously flawed.

As mentioned, the first is the assumption that humans are rational players in the great game of market roulette. They are not. Behavioural scientists have shown that while people are very good at recognising useful patterns and interpreting ambiguous or incomplete information in their decision-making, they are very poor when it comes to performing complex logical analysis, preferring to follow market leaders or flock according to the latest fashion. This can further amplify distorting trends.

The new theories of behavioural finance argue that during a bubble the rate of buying and selling can become manic, resulting in irrational decisions. Making money actually stimulates investor’s brain reward circuitry, causing them to ignore risk- increasing the difficulty of valuing stocks accurately.

But perhaps the most critically flawed assumption is that an economic system always reaches an ideal equilibrium of its own accord. In other words, the market is capable of benign self-regulation- automatically allocating resources and controlling excesses in an optimum way, best effected with minimum outside interference.

Since the nineteenth century the fundamental principle underpinning economics has therefore been based on the mythology that the economy is a system that moves from one equilibrium point to another, driven by shocks from external disruptions – whether technological, political, financial or cultural- but always eventually coming to rest in a natural equilibrium state.

The new emerging evolutionary paradigm however postulates that economies and markets, as well as the web, enterprises and the human brain, are all forms of complex systems in which agents dynamically interact, process information and adapt their behaviour to a constantly changing environment; never reaching a final stable equilibrium or goal.

In biological evolution, the natural environment selects those systems that are best able to adapt to its infinite variation. In economic evolution, the market is a combination of financial, logistical, cultural, organisational and government regulatory elements, which adapt to and in turn influence a constantly changing ecological, social and business environment.

In essence, economic and financial systems have been fundamentally misclassified. They are not perfect self-regulating systems. They are enormously complex adaptive networks, with topologies that include decision hubs, relationship connections and feedback loops linking multi-agent groups which interact dynamically in response to changes in their environment; not merely through simplified price setting mechanisms, tax and interest rate cuts, liquidity injections or job creation programs. They must be understood and managed at a far deeper level.

Modern evolutionary theorists believe that evolution is a universal phenomenon and that both economic and biological systems are subclasses of a more general and universal class of evolutionary systems. And if economics is an evolutionary system, then it follows there are also general evolutionary laws of economics, which must be understood and harnessed if it is to be effectively managed.

This contradicts much of the standard theory in economics developed over the past one hundred years.

The economic evolutionary ecosystem is now fed by trillions of transactions, interactions and non-linear feedback loops daily. It may in fact have become too complex and interdependent for economists and governments to control or even understand. Therefore, as several eminent complexity theorists have recently stated, it might be on the verge of chaos. Too much or not enough regulation can distort the outcomes further- creating ongoing speculative pricing bubbles or supply and demand distortions.

There is now an urgent need to understand at a much deeper level the genie that modern capitalism has engineered and released. This can only be done by admitting the current crumbling edifice is beyond repair and building a radical new model from the ground up; a system that incorporates the hard sciences of network, evolutionary, behavioural and complexity theory.

Tinkering around the edges with the old reactive tools is not an option anymore.
To have any real chance of harnessing the economic machine of the 21st century for the benefit of all human society, not just the wealthy, it must be modelled at the network level and managed autonomously according to adaptive evolutionary principles.

If a business as usual economic philosophy prevails, it is likely that the resulting ultra-massive waste of resources and social turmoil of a second GFC would be catastrophic for our civilisation.

Saturday, September 11, 2010

Future of Cyber-Infrastructure for World 2.0

Our future World 2.0 will face enormous challenges from now into the foreseeable future, including global warming, globalisation and social and business hyper-change.

Global Warming will create shortages of food and water and loss of critical ecosystems and species. It will require massive prioritisation and re-allocation of resources on a global scale.

Globalisation will require humans to live and work together cooperatively as one species on one planet- essential for our survival and finally eliminating the enormous destruction and loss of life that wars and conflict inevitably bring.

Social and Business Change will present myriad challenges relating to building and maintaining a cohesive social fabric to provide - democracy and justice, adequate levels of health and education, solutions to urban expansion, crime prevention, transport congestion and food and water security, in a fast changing global environment. This will require adaptation on a vast scale.

It is apparent that in order to meet these challenges, humans must harness the enormous advances in computing and communications technologies to achieve a complete makeover of the world’s Cyber-Infrastructure.

The infrastructure of the new cyber reality now affects every aspect of our civilisation. In tomorrow’s globalised world a dense mesh of super-networks will be required to service society’s needs- the ability to conduct government, business, education, health, research and development at the highest quality standard.

This infrastructure will be co-joined with the intelligent Internet/web, but will require additional innovation to facilitate its operation- a transparent and adaptable heterogeneous network of networks, interoperable at all levels of society.

In the last two decades tremendous progress has been made in the application of high-performance and distributed computer systems including complex software to manage and apply super-clusters, large scale grids, computational clouds and sensor-driven mobile systems. This will continue unabated, making the goal of providing ubiquitous and efficient computing on a worldwide scale possible.

But there’s a long road ahead. It is still difficult to combine multiple disparate systems to perform a single distributed application. Each cluster, grid and cloud provides its own set of access protocols, programming interfaces, security mechanisms and middleware to facilitate access to its resources. Attempting to combine multiple homogeneous software and hardware configurations in a seamless heterogeneous distributed system is still largely beyond our capability.

At the same time tomorrow’s World 2.0 enabling infrastructure, must also be designed to cope with sustainability and security issues.
It is estimated that The ICT industry contributes 2-3% of total Greenhouse Gas emissions, growing 6% per year compounded. If this trend continues, total emissions could triple by 2020. The next generation cyber-architecture therefore needs to be more power-adaptive. Coupled with machine learning this could achieve savings of up to 70 % of total ICT Greenhouse emissions by 2020.

But the world is also grappling with the possibility of cyber-warfare as well as increasingly sophisticated criminal hacking, with an estimated 100 foreign intelligence organisations trying to break into US networks. A global protocol safeguarding cyber privacy rights between nations, combined with greater predictive warning of rogue attacks, is critically needed. The next generation of cyber-infrastructure will therefore have to incorporate autonomous intelligence and resilience in the face of both these challenges.

To meet these targets a lot will ride on future advances in the field of Self-Aware Networks- SANs. Previous blogs have emphasised the emergence of the networked enterprise as the next stage in advanced decision-making. SANs are a key evolutionary step on the path to this goal. Self-aware networks can be wired, wireless or peer-to-peer, allowing individual nodes to discover the presence of other nodes and links as required- largely autonomously. Packets of information can be forwarded to any node without traditional network routing tables, based on reinforcement learning and smart routing algorithms, resulting in reduced response times, traffic densities, noise and energy consumption.
Another major shift towards a networked world has been the rise of Social Networks. These have attracted billions of users for networking applications such as Facebook, LinkedIn, Twitter etc. These are providing the early social glue for World 2.0, offering pervasive connectivity by processing and sharing multi-media content. Together with smart portable devices, they cater to the user’s every desire, through hundreds of thousands of web applications covering all aspects of social experience– entertainment, lifestyle, finance, health, news, reference and utility management etc.
With increased user mobility, location sharing and a desire to always be connected, there is a growing trend towards personalized networks where body, home, urban and vehicle sensory inputs will be linked in densely connected meshes to intermediate specialised networks supporting healthcare, shopping, banking etc.
The explosion of social networked communities is triggering new interest in collaborative systems in general. Recent research in network science has made a significant contribution to a more profound understanding of collaborative behaviour in business ecosystems. As discussed in previous posts, networked ‘swarm’ behaviour can demonstrate an increase in collective intelligence. Such collective synergy in complex self-organising systems allows ‘smarter’ problem solving as well as greater decision agility. By linking together in strategic and operational networks, enterprises can therefore achieve superior performance than was previously possible.
The key characteristics of the smart business network of the future will be its ability to react rapidly to emerging opportunities or threats, by selecting and linking appropriate business processes. Such networks will be capable of quickly and opportunistically connecting and disconnecting relationship nodes, establishing business rules for participating members on the basis of risk and reward.
This ‘on the fly’ capacity to reconfigure operational rules, will be a crucial dynamic governing the success of tomorrow’s enterprise. CIOs must also learn to span the architectural boundaries between their own networked organisation and the increasingly complex social and economic networked ecosystems in which their organisations are embedded.
In fact the business community is now struggling to keep up with the continuous rate of innovation demanded by its users. Social network solutions have the potential to help meet this demand by shaping the design of future architectures to provide better ways to secure distributed systems.
So what is the future of this new collaborative, densely configured networked world? What we are witnessing is the inter-weaving of a vast number of evolving and increasingly autonomous networks, binding our civilisation in a web of computational nodes and relational connections, spanning personal to global interactions.

By 2050 the new World 2.0 cyber-infrastructure will link most individuals, enterprises and communities on the planet. Each will have a role to play in our networked future, as the cells of our brain do- but it will be a future in which the sum of the connected whole will also be an active player.