Thursday, 6 October 2016

Complexity Science has a great Future

Dirk Helbing, ETH Zurich

Complexity Science is more important than ever. Even though there is now big data about everything in the world, I don’t believe that we don’t need science any longer, in contrast to what Chris Anderson has claimed in The End of Theory: the data deluge makes the scientific method obsolete (Wired Magazine 16(7), 2008). He basically suggested that if you just had enough data, the truth would reveal itself.

With the data available to us, is it possible to know everything and to build a crystal ball that allows us to see everything that is going on in the world in real time? In fact, such projects are under way, built by the military and research centers around the world. Stephen Wolfram has claimed that “Humans are more predictable than Elementary Particles”. And it seems that CERN has actually built such a prediction machine that uses artificial intelligence to learn patterns in the data of human social behavior.

But these are not just research projects, because "knowledge is power". So they are political projects, too. This raises the question if so much data will enable the ruling of a wise king or a benevolent dictator. Could we optimize the world? Could society be run like a giant machine?

There are companies that seem to be working on such concepts, such as IBM and Google. They aim at reprogramming our society and building an operating system for it that would guide our decision making, thinking, and behavior with personalized information.

This kind of technology has also become interesting for politics. We are heading towards remote control of people. This could be a powerful approach: Google could manipulate billions of people on our planet. It is, therefore, worrying is that people like Larry Page said that there are a lot of things he likes to do but unfortunately he cannot, because it is illegal. 

Our society is at a cross road. No question, we will live in a data based society – but what kind of society will it be? Feudalism 2.0, Facism 2.0, or Communism 2.0?

It is concerning that there are voices claiming that democracy is an outdated technology. We see that democracy is in trouble in some countries such as Poland, Turkey, and France. We might easily loose what we have built over hundreds of years – freedom, human dignity, fairness and justice, pluralism, democracy, participation, social norms and culture, security and peace, and many jobs. 

It is time to say Stop. The magic formula "more data = more knowledge = more power = more success" does not work in many cases. Correlation does not equal causation.

There is also a technical reason for this: even though processing power increases exponentially, data volume increases even faster – the fraction of data we can process is going down over time. Moreover, as we go on networking the world, systemic complexity is growing even faster, which implies a loss of top-down control and a need of distributed control.

We need to build a digital democracy, which we may call "democracy 2.0". We need to learn how to bring great ideas and the knowledge of many people and artificial intelligence systems together. For this, we need to build online deliberation platforms.

It is not the best individual solution that wins, but diversity: the combination of many solutions creates collective intelligence. Being confronted with so many problems in the 21st century, such as financial, economic and spending crisis, massive unemployment, responses of decision-makers have become ever more desperate.

Our main problem is the lack of sustainability. We are overusing the resources of the world, in particular nitrogen and phosphors, but also water – thus creating massive problems. 

We need a new kind of economic system, capitalism 2.0, which is liberal, democratic, participatory, social and ecological. This can now be built by combining the Internet of Things, block chain technology and complexity science. To build a circular economy, we need a system that can measure, value and trade externalities – external effects of interactions between people, companies and the environment. 

We have started to build such a system called "Nervousnet": a planetary nervous system based on Internet of Things technology run by the citizens. It is using smartphones as sensors to measure the environment. To be able to trust the system, informational self-determination is taken seriously. 

It also becomes possible now to map resources and who uses them with an app called But we need to go a step further. It is necessary to create a multi-dimensional incentive and reward system: "finance 4.0" or "social-ecological finance" that allows us to create feedback loops in the system in order to support favorable kinds of self-organization. 

This system can now be built. BitCoin has shown that it is possible to create money in a bottom up way. By measuring externalities of different kinds, people would create different kinds of money (and earn money), and this money would eventually rise to the top, thereby benefitting everyone. 

This would enable us to turn the digital dessert that Europe currently is into a digital rainforest with digital opportunities for everyone, where interoperability would allow to combine existing products and services in order to create new products and services, which unleashes combinatorial innovation. 

It is time to build this open and participatory information, innovation, production, and service ecosystem. Let’s do this together. 

Short Bio:
Dirk Helbing is Professor of Computational Social Science at ETH Zurich and member of the German Academy of Sciences "Leopoldina". Helbing is member of the Global Brain Institute and the International Centre for Earth Simulation. He heads the and initiatives, which want to open up the opportunities of Big Data and the Internet of Things for everyone, and leads the PhD program "Engineering Social Technologies for a Responsible Digital Future" at TU Delft. 

Sunday, 7 August 2016


by Dirk Helbing

Our innovation system has terribly failed. It is well designed to support gradual improvements of our knowledge and technologies. But it does not support disruptive innovations well, which would create new qualities and functionalities, or question the basis of our established knowledge and routines. Moreover, our knowledge does not keep up anymore with the pace at which our world changes, and solutions to new problems often come with serious delays. Therefore, we need to re-invent innovation. In particularly, we must learn to create systems embracing collective intelligence that surpasses the intelligence of even the brightest individual and of powerful supercomputing solutions. This cannot be based on top-down nor majority decisions. Diversity is absolutely crucial for collective intelligence to work…

The pdf of this article can be downloaded here

The innovation crisis

In times of economic recessions and political crises, innovations and new ideas are bitterly needed. But great ideas are rare, and many ideas that appear to be new just reproduce or re-invent what somebody else has thought before. As I will show below, it is very difficult to have just one or two great ideas in a lifetime that will survive for more than 50 years, or even change the world. This is bad, given today's world is changing faster than ever due to climate change, environmental change, demographic change, conflict, war etc. Do we innovate quickly enough? Does our knowledge still keep up with this rapid pace of change? I don't think so.

I certainly don't deny that the digital revolution is brewing a perfect storm. Within just a few years, we have seen many new technologies such as social media, Big Data, cloud computing, Artificial Intelligence, cognitive computing, Internet of Things, Blockchain technology as well as virtual and augmented reality – and I love them. But which of these inventions will create anything that will remain for 5000 years, or just 50? "Creative destruction" is often cited as ideal, because a new and better world order would be born from the chaos created. However, if we look back at the events so far, chaos has mainly born more chaos, in a gigantic global cascading effect that poses existential threats. Think of the events after September 11, 2001, or the financial, economic and public spending crisis, which hit us pretty unprepared.

In 2013, The Economist started "The great innovation debate".[1] It raised the question, whether we will ever invent something as important as the toilet again. It is often said that Big Data is the "oil of the 21st century", but people increasingly add that, apparently, we haven't invented the motor yet to use it. Or if we have invented it, we have failed to build it due to political or economic constraints. Let me stress that I love "moonshot projects", including many of Google's efforts in this direction. Some people believe that, thanks to superintelligent systems, we will have all problems of the world solved by 2036 - apart from climate change.[2] However, this seriously underestimates the nature of complex systems such as our society and economy. For sure, we will be faced with new challenges such as cyber threats. And so far, none of our attempts have been able to restart the engine of the world economy, or create prosperity for all - a fact that has been criticized by an Open Letter on the Digital Economy, which obtained a broad support.[3] In the last decades, inequality has further increased. In times of stagnation, this means that the lower and middle class, in other words: the small and medium sized companies had to pay for this. The diverse "ecosystem of customers and firms" that makes up a thriving economy[4] has increasingly degraded. It trends towards the creation of monopolies, particularly in the IT industry. While these monopolies increasingly claim global leadership where our societies should head, I don't see that they would have the recipes to create global well-being.

According to mainstream economics, a lack of innovations should never occur. If a problem would just grow big enough, they argue, there would be an increasing willingness to pay a high price for a solution. Accordingly, scientists and engineers would be incentivized to work hard to find a solution. Therefore, any big enough problem would be fixed sooner or later. However, even though the fact of global warming due to our carbon-based economy has been known already in the 1960ies,[5] fifty years later we are still lacking a solution to the problem. Climate change may erase as much as one sixth of all species, and it poses an existential threat to humanity. By signing the Paris Climate Agreements, it has finally been admitted that innovations did not manage to solve this problem, even though many billions have been invested.

I personally believe those innovations haven't gone deep enough. We need to think out of the box, but we have stayed within it. As Albert Einstein said, we can't solve problems with the same kind of thinking that created them. Let me illustrate this for the example of the future prospects of our world. Back in the 1970ies, The Club of Rome published their Limits to Growth study. No matter how hard they tried, they could not find a sustainable development path for the Earth. We would, therefore, run into economic and population collapse. Even though the study was highly controversial, the Global 2000 report commissioned by president Bill Clinton came to very similar results. So, if we want to avoid economic and population collapse, we must change the system of equations underlying the simulation scenarios. This means nothing else than the need to change our socio-economic framework. That is exactly what "finance 4.0", a new socio-ecological financial system could do – by measuring, valuating, and trading externalities and creating a multi-dimensional incentive system that would boost a circular and sharing economy by unleashing powerful market forces.[6] It is time that responsible economic and political players take action on this.

And what about our health system? In fact, the pharmaceutical industry is in big trouble, too. Most of the relevant companies have a declining number of new drug registrations. Even though the large multi-national companies are buying lots of startups to stay on top of innovation, this hasn't changed their situation much. In the meantime, we are running out of antibiotics that are effective against multi-resistant strains of bacteria. This is a serious issue, indicating a failure of our current research and development approach.

Some people are even more pessimistic than that. They claim that there haven't been any great innovations since Darwin's theory of evolution and Einstein's theory of relativity. Unfortunately, this is probably true, and it has reasons. Science is increasingly run like a business, measured by performance indicators. But while we perform better and better according to these indicators and despite the highest number of publications and patents ever, many of the problems our society is facing haven't been fixed. We are far from having the answers that our quickly changing world demands from us. For example, many people think we are far behind the Millennium goals or the United Nation's 2050 development goals.

Unfortunately, the need to accelerate innovation will increase even more. According to Moore's law, computational power is doubling every 18 months. In 10 to 20 years, supercomputers are expected to exceed the processing capacity of a human brain. We will have computer programs capable of teaching themselves, robots producing other robots, and they will quickly improve over time. Hence, computer algorithms and robots will take over many of today's jobs. Anything that follows certain procedures, probably around 50 percent of all jobs in the industrial and services sector, could be performed by them cheaper and better. How can we cope with this challenge of having to reinvent half of our economy in just two decades? And how can we adapt our social, economic and legal system over such a short time? We need an Innovation Accelerator. But how would it work?

An outdated innovation system

Let us first analyze how we innovate today. While I will focus on the academic system, I expect that similar problems occur in industrial innovation systems as well. Currently, innovation is mainly happening in a competitive way. Each scientist, each company competes against all the others. Such innovation is expensive, slow, costly, and duplicates many results. This has led to an increasing percentage of programmatic research, which works roughly as follows: First, a ministry or agency determines research needs, emerging trends and knowledge gaps. It probably takes a few years until these become obvious, and it also takes some time to mobilize the budget (see, for example, the European Union's famous 7 year plans). But once a problem has been identified and the budget set aside, a call for proposals is launched.

These calls are usually oversubscribed, because there are never enough resources for everyone. This problem is "solved" by so-called "scientific beauty contests." Basically, one puts as many obstacles in the way as needed to retain the number of proposals that can be funded, i.e. one makes proposal-writing a complicated and time-consuming task. Selecting proposals is an equally complicated task. It is usually based on peer review – a process that consumes an increasing fraction of time as well. In the meantime, scientists probably spend about 40 percent of their time on proposal writing, reporting and reviews. Obviously, this time and money is lost for research, but the reason given for the inefficient administration is "having to justify how tax payers' money is spent".

Forget about determining the best innovations beforehand!

Can we at least be sure that the best contributions are selected? This is hard to say, as most non-funded projects are never carried out. However, the same peer review process is applied by scientific journals to select manuscripts for publication. Since many manuscripts that are rejected in the originally chosen journal are eventually published in a lower-ranked journal, we know a bit more about the quality of rejected as compared to accepted manuscripts. Surprisingly it turns out that a significant percentage of rejected manuscripts performs better than the ones that were accepted in the journal of choice.[7] A further surprise is that the majority of accepted and published papers performs significantly below the average of the journal, because a few of its publications generate most of its impact. This basically shows that quality is very difficult to judge. Quality may become obvious only over a long time. In fact, the recommendations of referees are often extremely divergent, particularly for innovative contributions, for which established standards do not exist.

But let's assume the best proposals were selected after a typically half-year-long review process. Then, one must find suitable staff to work on the project. The working contracts will start about 6 months after the acceptance of the proposal. The project will typically take 3 or 4 years, basically until a PhD is obtained. Publication of the research results will require between 6 months and 3 years, depending on the research field and journal. In any case, it's safe to say that this is not a fast process, and that it may easily take 10 years between the emergence of a new problem and its solution. If such a solution finally enters the knowledge core of the field (i.e. when it enters educational books and programs, which is the exception rather than the rule), it will take another 10 to 30 years, until it becomes best practice in business and administration. In cases of commercial industrial bias, scientific progress is often delayed by at least another 20 years, as it happened in the tobacco and energy sectors.[8]

This situation is reinforced by making science increasingly dependent on industrial funding – typically in order to fix today's problems rather than thinking ahead to find new approaches for the future. The only exception are projects of strategic nature that are of national importance. Many of these are aimed at accumulating power. But power is not the solution to many of our problems in a highly networked world, where strong interference will have unexpected side effects, feedback effects and cascading effects and often destroy structures that are essential for our society and economy to function.

Therefore, my personal judgment is that, on the one hand, our research and development (R&D) system has become increasingly dysfunctional. On the other hand, I know that many of the most successful publications are produced under difficult circumstances – they result from spontaneous ideas that someone decides to follow in the spare time, even though there is basically no funding for this. I have often wondered, why this is the case. To get funding is not the biggest problem. The problem is that we are asked to plan innovations, while the best ideas just happen. Funding institutions love well-elaborated proposals, but once one can elaborate new ideas in detail, they are not anymore cutting edge.

Great ideas must be pursued immediately, without the delays imposed by conventional funding mechanisms. However, our current innovation system makes scientists spend their time on ideas they get money for, and usually there is no time left for others. This effectively undermines academic freedom, and in many cases, certain innovations will be delayed for years, or even buried forever. By the time money becomes available, there will be other exciting ideas.

Thus, the crucial question is how to produce results more or less in real-time? Given that, today, it takes about 30 years from an invention to the real-world application of ideas, and many good ideas will never make it - how can we shorten this process to 5 years, or even 5 months or 5 weeks? And how can we increase the success rate of inventions? If we understood this, we could produce a dramatic innovation boost. Given that we have already more than 20 million unemployed people in Europe alone, such an innovation boost would be bitterly needed.

Everyone wants innovations, but opposes them!

To improve the innovation mechanism, it is important to first understand the nature of innovation a bit better. It turns out that there are two different kinds of innovations: gradual and disruptive innovations. Gradual innovations can be measured according to established standards of a field. They may best be characterized as "improvements" – such as a motor that consumes less energy and produces less emissions. Here, it seems reasonable to expect consensus of the reviewers in a funding board.

Pioneering research, in contrast, produces disruptive innovations, exploring or creating entirely new quality dimensions. These are the "true" innovations, which decision-makers and business people are usually keen on. Thus, why don't they happen more often? Almost by definition, disruptive innovations can't be assessed with established standards. They transcend existing categories and require one to think “out of the box.” Consequently, such innovations are often highly controversial, and majority decisions of the reviewers in a funding board will rarely support them.

History shows that basically every disruptive innovation has been opposed in the beginning. The following quotes are quite illustrative. Ten years after the first successful test of electric light bulbs on October 22, 1879, Thomas Edison said: "Fooling around with alternating current is just a waste of time. Nobody will use it, ever." But today, everyone is using this kind of electricity 24 hours, 7 days a week. Or take the US president Rutherford B. Hayes. After a demonstration of Alexander Bell's telephone in 1872, he concluded: "It's a great invention but who would want to use it anyway?" Later, the inventor Lee De Forest (1873-1961) stated: "While theoretically and technically television may be feasible, commercially and financially it is an impossibility." Similar opinions were voiced, when the radio, planes, drilling for oil, or nuclear energy production were proposed.

It's no wonder that Alexander von Humboldt (1769-1859), one of the great discoverers of the world (and inventor of our modern university system) came to conclude: First, people deny that the innovation is required. Then, people deny that the innovation is effective. Afterwards, people deny that the innovation is important, and it will justify the effort to adopt it. Finally, people accept and adopt the innovation, enjoy its benefits, attribute it to people other than the innovator, and deny the existence of the previous stages. In other words: most innovations won't make it.

The famous quantum physicist Max Planck (1858-1947) even claimed: "Science advances one funeral at a time." This is mainly a result of the "rich gets richer effect," as Robert Merton (1910-2003) called it: while new inventions are made all the time, highly referenced work tends to get an even increasing amount of attention. This creates a threshold effect: only ideas that manage to get above the attention threshold will have a chance to win through. Such ideas are called "revolutionary ideas" and cause sudden, fundamental changes of our understanding or even of our world, so-called paradigm shifts, as analyzed by science historian Thomas Kuhn (1922-1996).

Revolutionary breakthroughs trigger an avalanche of new ideas, change the perspective of our world, and have the potential to transform our reality. One spectacular example is the replacement of the human-centered ("geocentric") worldview assuming the Earth to be the center of the universe by our current view that our planets would circle around the sun ("heliocentric" view). This shift goes back to observations of Nicolaus Copernicus (1473-1543) and theoretical work of Galileo Galilei (1564-1642). Later, it allowed Isaac Newton (1642-1726) to come up with his equations for the dynamics of celestial bodies. Without these discoveries, we would not be able to send rockets to the moon or have satellites circle around the earth. What seems to be a natural point of view today questioned the Christian worldview so much that Galilei was sent to prison. Only 350 years later, the Catholic church apologized for this.

The discoveries of Charles Darwin (1809-1882) were not less shocking. His theory of evolution – implying that humans were descendants from apes – largely replaced the idea of divine creationism, and is still questioned by some people today. However, without this paradigm change, it would be hard to imagine genetic engineering today. Or think of the theory of relativity by Albert Einstein (1879-1955). Without it, we would not understand how to produce nuclear energy or how to operate the Global Positioning System (GPS) exactly. In 1931, a provocative book entitled "100 Authors Against Einstein" even tried to discredit his work. Nevertheless, after a couple of decades, Einstein's counterintuitive predictions were finally confirmed. Such revolutionary ideas are extreme events – one might even say: "black swans". They occur only every fifty or hundred years or so.

But if great ideas cannot be identified beforehand, why don't we engage in refunding excellent work that has already happened, rather than in funding people for impressive promises made in lengthy and complicated project proposals? In other words, why do we pay money for the best promises, and not the best results? Wouldn't it save billions of tax payers' money, if the relatively few brilliant minds that exist could concentrate on innovations rather than on proposal writing, evaluation, and reporting? The good point about the scarcity of ground-breaking ideas is that funding agencies would have more than enough money to (re-)fund them. The open problem though is how to identify them (which, as I said before, cannot be well done by consensus or majority decisions of funding boards, as long as the ideas are young and quality criteria and research communities are not yet established).

Detecting game-changing ideas and the innovators behind them

But there is, in fact, a way of detecting where new ideas are produced, and where they are consumed. I did such a study together with Amin Mazloumian, Katy Börner, and others.[9] Each scientific publication refers to others it has been inspired by. Therefore, one can identify the flow of ideas in the world, and what are the places that produce ideas that are over-proportionally successful (shown in green in the figure below).

It is even possible to reveal what are the main ideas discussed in these publications – by analyzing the spreading of "memes." Memes are single words or combinations of words that appear in texts such as scientific publications. In physics, "atom" or "quantum mechanics" or "high-temperature superconductivity" would be such examples. In fact, my postdoc Tobias Kuhn and I, together with Matjaz Perc, have scanned the abstracts of all publications of the American Physical Society for such memes. This allowed us to identify the first occurrences of any new word or combination of words (meme). Of course, scientists come up with new terms all the time, but only in a few cases is the usage frequency quickly growing significantly in time. If this is the case, a new trend is born. Moreover, there is another property that characterizes Earth-shaking ideas. Their memes are "inherited" through the citation graph, i.e. they spread through mentions in later publications of colleagues. This separates important scientific concepts from meaningless memes. In fact, the history of the most important fields in physics can be determined in a fully automated way[10] (see figure below).

Note, however, that data mining of publications and citations cannot only identify new trends, but also the key researchers in the field. Together with Amin Mazloumian, Santo Fortunato and others, I found that a milestone paper is one that doesn't only attract many citations, but also draws more attention to the entire body of work of a researcher. Consequently, many of his or her previously published papers will be cited more frequently. In other words, a milestone paper boosts other papers of the same author, which can easily be measured.[11] This fact allows one to identify rising stars at the firmament of science, which usually stay on top of their respective fields in terms of citations for their entire careers. In fact, all Nobel prize winners have such a milestone paper, which boosted the visibility of their whole body of work and, thereby, their entire career. But there are only a few scientists, who have two or more big boosts in their careers, i.e. two or more milestone papers (see figure below). Hence, it is really affordable for a research institution to flood the few researchers, who succeeded in having a relevant boost effect, with money.

Chance – the true father of new ideas?

However, the perhaps most surprising implication of the above discussion is that it is often impossible to determine in advance how successful an idea or someone will be. Scientists often have difficulties to judge themselves, which of their papers will be most successful. And this reveals another important feature of the nature of innovation. A successful innovation requires at least two things: First, an invention must be made, and second, it must spread. We often believe that great inventions will result, if well-skilled experts undertake a large enough and systematic effort. However, they rarely occur without the contribution of chance. You may call it "luck" or "serendipity" – naming it "creativity" certainly sounds more deserving. In fact, many great inventions are based on trial and error, and generating suitable settings, which allow randomness to contribute, is an art. For example, the famous inventor and electrical pioneer Thomas Alva Edison (1847-1931) noted: "I have not failed. I've just found 10,000 ways that won't work." Moreover, even people who made a great invention, often do not recognize its potential. For example, James Clerk Maxwell (1831-1879), who formulated the theory of electricity once said: "I do not know what electricity is good for, but I'm pretty sure that Her Majesty's government will soon tax it." How right he was! But practical applications spread much later. The same was true for the MASER (the predecessor of the LASER). When the invention was made, it appeared pretty useless, but in the meantime, a huge market emerged around LASERs.

Social factors matter

The idea that a successful innovator is a lonely genius is often misleading. Successful innovations are also the result of social processes. The ability of innovations to spread is a crucial aspect. In fact, the "ignored genius" has become a well-known proverb. Sometimes, ideas are born ahead of time. But in order to spread, they must come at the right moment. The context must be fitting. It's helpful if many people have already been waiting for the idea to occur. So, what determines whether an idea comes just at the right moment? This actually depends a lot on the social network. Looking into the origin of citations, which are often used to quantify the success of ideas, my PhD student Christian Schulz made an important discovery: The first citations are usually self-citations, then citations by co-authors tend to follow, and later citations of co-authors of co-authors. Only less than 40 percent of citations on average come from people not related to the inventor directly or indirectly through the co-authorship network. Hence, one may interpret the co-authorship network as the "infrastructure" through which ideas spread. In other words, an isolated inventor will rarely succeed. It's important to have a social network through which the ideas can successfully spread.

To spread an idea, one must furthermore "get it right." This is not just a matter of formulating the idea correctly – it's also a matter of presentation, i.e. the storyline or narrative. An idea that fits current trends, but creates the right degree of intellectual stimulation – not too much and not too little – is best. If the idea is too innovative, most people will not understand it, and so it will not spread.

It is also important to recognize that an initial idea is almost never right. But if you don't dare to expose it, you will never get it right. Hence, another important success factor is to have a social environment that gives critical feedback. I often say that I owe my competitors and enemies the most. While critique is annoying, it helps one to frame an idea well, and it often reveals the unsolved problems that others are struggling with, i.e. the solutions the world is waiting for. Why is this important? Because it means that innovations need an environment that allows one to make mistakes. However, today's research and development settings rarely provide such an environment, and scientific consensus is favored over controversial debates. The Silicon Valley is rather the exception than the rule. It's known to allow people to experiment and make mistakes. Those who fail, are not singled out. They will find venture capital and another job again, because they are considered to be people, who have (probably) learned a lesson. Some people think, this particular culture has a historical reason: the birth of the hippie culture in the 60ies. It seems that true innovation needs rebels, who are ready to challenge the establishment.

But how can we unleash the power of new ideas today? We need to give scientists more freedoms. Instead of asking for proposals, I suggest that interdisciplinary panels should try to determine the smartest junior scientists based on their CVs and presentation of ideas, giving them a safe salary for at least four years. For the sake of collaboration and support, the successful candidates would then look for a suitable academic team to realize the ideas they have in mind. The corresponding professor or leader of a research team could either accept or decline the applicants. In case of acceptance, the team of the host would get some overheads. I call this approach the “marriage principle,” and I believe it could largely replace the proposal-based funding system, which slows everything down. It would create much better chances for new ideas compared to those established earlier on. For sure, such a bottom-up approach, empowering junior scientists, would be more responsive to emerging challenges than the research and development system we have today.

Re-inventing innovation

But there are at least two more things that can help to accelerate innovation dramatically: the way we collaborate and the way we handle intellectual property rights (IPR). I will argue that, to catch up with the pace at which our world is changing, the current IPR protection approach is a great obstacle, and that we need a novel co-creation paradigm. In fact, using information and communication systems, we can re-invent innovation. Currently, people don't like to share their best ideas, because they prefer to be rewarded for them rather than allowing other people to become successful or rich on them. Therefore, it often takes years until an idea is shared with the world through a publication or patent. What, if we instead innovated globally and cooperatively from the very first moment? Say, an idea is born in America, and it is shared through a public portal such as github. Then, experts from Asia could work on these ideas just hours later, and those from Europe would build on their results. In such a way, one can create a new research and development paradigm that never sleeps; one that overcomes the limits of a single team; one that embraces "collective intelligence" by integrating diverse perspectives.[12] Such an approach actually produces considerable synergy effects. As my colleagues Didier Sornette and Thomas Maillart have recently shown: 1+1=2.5. Specifically, the collaboration of two people on producing open source software creates outcomes that would otherwise take 2.5 people. Already before, Geoffrey West, Luis Bettencourt, some others and I have found evidence for a similar scaling law for cities:[13] productivity that depends on social interactions grows super-linearly with city size (namely according to a power law with an exponent around 1.2). This is probably the main reason for the dramatic on-going urbanization of the world.

Now, with Internet forums of all kinds, something like virtual cities have grown. Many citizen science projects (and also the famous Polymath project on collaborative mathematics[14]) underline that a crowd-based approach can outperform classical approaches in research and development. So, given the great advantages of collaboration, what are the main obstacles to the immediate sharing of ideas? I would say, mainly the lack of proper incentive systems. Researchers live on two kinds of rewards: their limited salary and the applause they get in terms of citations, i.e. the mentions they receive by fellow scientists. Many of them would not like to share their best ideas with others, before they have published them, nor would they like a company walk away with the idea and make a lot of money on it without sharing the profits in a fair way. Patents and other intellectual property rights are intended to protect the commercial value of ideas and thereby to stimulate innovation. However, in the area of digital products, patents seem to be more an obstacle to innovation rather than a catalyst for it.

Patents on ideas are a bit as if everyone would own a certain number of words and would charge others for using them – this would certainly considerably obstruct the exchange of ideas. In fact, it has recently been difficult to enforce hardware and software patents, and we see ever more patent deals between competing companies. It has also been found that opportunities to watch music videos free of charge can promote their overall sales. Interestingly, Elon Musk leading the electro-car company Tesla Motors has recently decided to allow others to use their patents. Furthermore, Google TensorFlow and OpenAI have made their artificial intelligence algorithms open source. All this might indicate that a paradigm shift regarding Intellectual Property Rights in favor of Open Innovation is just around the corner.

Micropayments are better than protecting intellectual property

So, why not pursue an entirely different intellectual property rights (IPR) approach – perhaps as replacement, perhaps in parallel to the intellectual property approach of today? At the moment, we are trying to prevent the copying of digital products and handle them as if they were material ones. However, it is the nature of information that it can be reproduced as often as we like and almost for free. Information is a virtually unlimited resource. In contrast to material resources, it would allow us to create benefits and opportunities for all, thereby overcoming poverty and conflict. Turning information into a scarce resource and protecting it from duplication is stupid. It is against the "nature" of information. Therefore, what if we allowed copying, but introduced a micropayment system that ensures that every copy generates some profit for the originator? Under such circumstances, we would love duplication!

Rather than complaining about copies, we should make it easy to be paid for the results of creative and innovative activity. Remember that, some time back, Apple's iTunes made it simple to buy songs and download them, for 99 cents each, thereby overcoming the need for individual negotiations. It would be great to have a similarly simple, automatic compensation scheme for digital products, ideas and innovations. Today's powerful text mining and pattern recognition algorithms could be the basis. Then, whenever another person's or company's idea would be used, there would be an automatic payment, which could depend on the amount of investment made, the novelty of the invention, the advance (the innovativeness), and the "age" of the innovation. This would overcome obstacles like patents, and it would encourage cooperative innovation activities without having to worry that someone could steal an idea.

In perspective, competition would be replaced by co-opetition, i.e. a combination of competition and collaboration. Co-opetition would accelerate innovation enormously. In fact, an Innovation Accelerator should make it simple for people with compatible interests to find together and join their diverse skills in order to carry out an interesting project together. It is rather unrealistic to expect that a scientist who does fundamental research would eventually do applied research, and then develop a product and establish an own spin-off company. This does not happen very often. It would be much better to create alliances between Universities and Universities of Applied Sciences and start-ups or companies that take care of the different steps of this maturation process, ideally all on one innovation campus. I would even say, we need entirely new R&D and education environments for digital innovation, as we have special research and education systems for agriculture, for engineering (the industrial society), and for business (the service society). Such innovation and education environments should make use of the opportunities created by modern platforms such as edx, Khan Academy, Udacity, Mendeley or ResearchGate, but also of Amazon Mechanical Turk, Innocentive and other crowd sourcing, crowd funding and citizen science platforms. To ensure an efficient transfer of knowledge, it is now important to create an innovation ecosystem,[15] where the best of all knowledge and ideas can come together. What's the big idea that you will contribute to it?

Further reading:

D. Helbing and S. Balietti (2011) How to create an innovation accelerator. EPJ Special Topics 195, 101–136.

F. van Harmelen, G. Kampis, K. Börner, P. van den Besselaar, E. Schultes, C. Goble, P. Groth, B. Mons, S. Anderson, S. Decker, C. Hayes, T. Buecheler and D. Helbing (2012) Theoretical and technological building blocks for an innovation accelerator. EPJ Special Topics 214, 183–214.

J. Johnson, S. Buckingham Shum, A. Willis, S. Bishop, T. Zamenopoulos, S. Swithenby, R. MacKay, Y. Merali, A. Lorincz, C. Costea, P. Bourgine, J. Louçã, A. Kapenieks, P. Kelley, S. Caird, J. Bromley, R. Deakin Crick, C. Goldspink, P. Collet, A. Carbone and D. Helbing (2012) The FuturICT education accelerator. EPJ Special Topics 214, 215–243.

S. Balietti, R. Goldstone, and D. Helbing, Peer review and competition in the Art Exhibition Game (2016) Proceedings of the National Academy of Sciences of the USA 113, 8414-8419.

M. Schich, C. Song, Y. Y. Ahn, A. Mirsky, M. Martino, A.L. Barabási, and D. Helbing (2014) A network framework of cultural history. Science 345, 558-562.

D. Helbing, The Automation of Society Is Next: How to Survive the Digital Revolution (CreateSpace, 2015).


[2] and cybercrime, that's what I learned from Jim Spohrer of IBM.



[6] D. Helbing, A Digital World to Thrive in:; Why We Need Democracy 2.0 and Capitalism 2.0 to Survive:; Society 4.0: Upgrading society, but how?

[7] In fact, a number of Nobel-prize winning discoveries have been rejected by scientific journals, or published in low-level journals.

[9] A. Mazloumian, D. Helbing, S. Lozano, R. P. Light and K. Börner (2013) Global multi-level analysis of the 'Scientific Food Web'. Scientific Reports 3, 1167.

[10] T. Kuhn, M. Perc, and D. Helbing (2014) Inheritance patterns in citation networks reveal scientific memes. Physical Review X 4, 041036.

[11] A. Mazloumian, Y-H. Eom, D. Helbing, S. Lozano, and S. Fortunato (2011) How citation boosts promote scientific paradigm shifts and Nobel Prizes. PLoS ONE 6(5), e18975.

[12] A.W. Wooley et al. (2010) Evidence for a collective intelligence factor in the performance of human groups. Science 330, 886-888; S.E. Page (2008) The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies (Princeton University); Networked minds: Where human evolution is heading,; D. Helbing and S. Klauser, How to make democracy work in the digital age,

[13] L.M.A. Bettencourt et al. (2007) Growth, innovation, scaling and the pace of life in cities. Proceedings of the National Academy of Sciences of the USA 104, 7301-7306.

[14] T. Gowers and M. Nielsen (2009) Massively collaborative mathematics. Nature 461, 879-881.

[15] D. Helbing, Distributed collective intelligence: The network of ideas,