The psychology behind Internet of Behaviors (IoB)

Not so long time ago, only a fool or an artist could ask and expect a real answer to the question ”What is happening to our world?” Today’s knowledge companies are already making mountains of money by answering it. But reading the Big Data optimists’ claim that we can soon predict and recognize human behaviors I can see Sigmund Freud lurking behind the hunger for ‘objective truth’ about our behaviors (its 2100 century version is called ‘unedited behavior’): http://blogs.hbr.org/cs/2012/09/predicting_customers_unedited_behavior.html?cm_mmc=SocialHub-_-3271-_–_-9135650988269047185 It is time to turn to the next question “What are we intending to do to the world?”. There is no ‘IntentionGoogle’ to satisfy that curiosity.

As an example, at the writing of this, my blog has had about 5000 visits, not a big number in today’s blogsphere but I would really like to know what were some of the true intentions of my readers when they met my texts. Knowing that I would understand a lot more about the world and in a way that could teach and touch me.  Current technologies and apps do not offer that luxury to us.  What a shame.

So, what follows is the second part on ‘The Internet of behaviors (IoB)’  text https://gotepoem.wordpress.com/2012/03/16/internet-of-behaviors-ib/ and idea where I now try to explain the human background of it and describe its potential value in improving our ability to see intentions and to know better what is about to happen in the connected world.  IoB is an easy technical concept but psychologically a very complex one. It has many risks, but offers also huge potential if implemented in a right way.

Blind to intentions, blind to the future

It is for personal pleasure and benefit that we want to know the intentions of our trusted friends and fellow citizens. In everyday life we simply ask them. We do this at individual and organization level and use the acquired intention knowledge to prepare our near or far future. Parents, company executives and medical doctors mine intention data routinely from children, employees, and patients. All this happens automatically, but we do not call it ‘intention knowledge’, and actually, I have never even seen this term used before.

Sociologists and politicians teach us that the dominant powers in the world are structural and ideological in nature and that individual behaviors are but direct reflections of these structural forces. Accordingly, individual intentions are just buried noise in these sociological masses. Psychologists on the other hand, are  trapped with their academic and context independent personality and intelligence measures that actually have almost no predictive value in life. https://gotepoem.wordpress.com/2011/10/13/entropy-of-the-human-soul/

It is not common knowledge that, for example, the best known personality factors “The big five” have close to zero % explanatory value in predicting future success in work performance. Knowing all this it is no surprise that the futures researchers with their forecasting methods have problems in dealing with the future of human and socially driven technology developments cf. e.g. http://www.meaningprocessing.com/personalPages/tuomi/articles/ForesightInAnUnpredictableWorld-extract.pdf.

It is time to learn to deal with the intentions and to accept that several drivers exist simultaneously behind it.  Intentions are true subjective phenomena.  It was almost impossible to understand what made the Norwegian murderer to act as he did – there is no one explaining variable, not even a complex set of them like intelligence, mania, political view, or personality, but he had intentions. He had clear and explicit intentions that integrated all his psychological powers. This made him dangerous.

In the connected world the individual has new potential in good and bad: the acts of creative souls and stubborn maniacs have repeatedly surprised us; first with their intentions – which typically have been recognized late like in the case of Steve Jobs – and then by their revolutionary or transformational actions that have changed our world and its structures.

Intention is more than wishful thinking and self-deception

What if we could know more about the intentions of individuals, organizations, or even about artificial systems? Wise parents have always been sensitive to the acceptable or questionable intentions of their children, just like civilized managers and coaches have been in order to interfere or provide support early. Perception of intentions requires true intelligence.

The core idea behind the Internet of Behaviors (IoB) is to provide means to personally code, register, share, and address (individual or organizational) behaviors and to use pre-defined IP (v6) addresses to be used as individuals or any communities see as best.  By IP a person can indicate his current intentions and behaviors. In doing this, it is not necessary to think about ‘addresses’ – it is possible to use seamless or ubiquitous future UIs of which various designs can be easily imagined.  Thinking about the potential of the future technologies it is a fascinating challenge to design such interfaces for individuals, friends, families, couples, colleagues, and to trusted outsiders. What could a personal repertoire of IPs be like, and what can it offer?

Random behavior addressing by data giants

Our behaviors are already being addressed by the knowledge giants. They have ways to represent our expected behaviors (traveling, shopping, work, hobbies) but it is more like shotgun addressing with a very coarse aim. Actually they do not call it ‘intention knowledge’ although that is exactly what it is, only that it is guessing because we have not been asked.  All this happens invisibly and typically we are never informed about the data structures that point their finger at us. The knowledge businesses look at our economical histories, our social and national background, education, habits, networks, locations and then they combine and pack these data to sell it further. They have no idea of our intentions. But they behave as if they were the owners of our intention knowledge because we have no means to manage it.

Not unlike the respected giants also pirates use our data without our knowing of it. But they really don’t know what we actually want, what we intend to do and why we do things our own way. Most of their data is inferred (relative to mass profiles) and it is only superficially based on what we have directly told or otherwise expressed. They cannot avoid the human inverse problem: as individuals and even larger communities we do apparently same things for completely different reasons.  The idea behind IoB is to make this behavior addressing visible and to declare it our own property.

Expressing ourselves in media

Despite the great social apps like fb, Linkedin, Pinterest, Goodreads, and many newcomers it is astonishing how impotent the current information environment is in offering us means to express ourselves, in what we are doing and engaged with and what we intend to do. A typical way to start a conversation over a mp “Are you busy?” We have no idea what our contact might be up to during the phone call.  Similarly, most network apps collect superficial and static information about us, probably because they suffer from the history of a questionnaire mania and have remained elated by the modern and easy data collection tools. As a result, they lag the real technological potential to support our opportunities to make us visible in a way that is beneficial to us and to our friends and trusted ones. Simultaneously there are numerous devices on the market that record your hear rate, sweating, running and whatever and they do not tell much more – but they could.

Future library – a service example

Knowing that a person is in the library, especially in a modern one, does not tell much why he is there. No present library offers a service, by which you would have everything you need ready for you, from books to your relevant knowledge sites, open documents and the work space that you need when you enter the library. You could have your last manuscripts exactly in the same state as you last left it and perhaps even announcements open for you about potentially interesting new information or comments to it if you happen to be writing one. In the old days it was easy to accomplish this at home and just leave books open and the manuscript beside them. Now we have learned to manage the work space configuration every time we start working on something.

Why don’t we have such services? The reason is simple: the library does not know you and it cannot infer your individual reasons to be there. And it never asks. This is strange because often you just tell your friends or colleagues:  “I’m going to the library to study  X”, indicating a real and simple intentional act. One has to be a paranoid to claim (still many claim this) that it is not possible to trust what people say in such situations. But we tell it because it is useful information for our co-existence. We all know why it is useful in a number of ways; even monkeys signal their intentions to each other and reindeers have systems to share intention knowledge when they move in flocks. (For a fascinating lesson about reindeer ‘team behavior’ by Esa Kirkkopelto, see http://www.toisissatiloissa.net/esitteet/PORO_valmis_press.pdf .)

Why should we let others know our intentions? Having a means to express the intentions – before we actually do anything – is a most valuable and inspiring form of knowledge. Between lovers it is a most fascinating form of knowledge but the library has another kind of a role. A library visit is a repeating behavior pattern that is easy to express by opening and activating the IB address assigned to this specific behavior. Anyone with the access rights could then make use of that knowledge.

Intention occurs before action and it allows intelligent preparations. This is where Big Data analysis lags behind: it is always late in knowing (or guessing) our present individual intentions. The receiver who gets this knowledge from the person has an advantage over any Big Data analysis outcome.  There is no limit to the richness of the intention knowledge itself and to the way it is expressed – only imagination sets its limits.

With the simple message, and a unique IoB address related to it  ‘I’m going to study and work on X’, a place (library) and time tag, and the student could indicate that he is a client that is ready to enjoy any possible service and data that is targeted to this address. If he happened to live far away public transport info could be of help. On the other hand, when he would express this same intention to ‘study and work on X’ at home, in his own ict environment the same service could be available to continue exactly where he last time left, in the texts, with open documents and net connections.

For this type of services to work a permission arrangement is necessary between the IoB address owner and the potential service provider or anyone else enjoying the right to access. In many cases it is not difficult to achieve. For example, in the library case, often libraries require a library pass and in giving it they also record relevant information from the new customer.  Many other coordination, security-checking and practical arrangements can be easily imagined.

When we share the intention information about the meaning of the address with someone else (like the grocery store, car sales companies) we give a permission to be approached or followed by these trusted parties. They, on the other hand can either collect information or provide data or services to the person. As a result communication will be coordinated with our relevant intentions and behaviors.

Scaling up with finances – shaking the banks

Putting the security aspects aside first, what if all bank customers of a bank, say in Spain, had a quick and simple means to express how they intend to behave with their banks, for example, how much cash they intend to take out from their accounts over the next 6 months? This could be a two-way arrangement: the bank could own a large pool of IoB addresses that it can offer to its clients to use for their specific purposes. Each of its customers then has an individual address sub-pool that allows a number of expressions (willingly) to indicate his variable economical actions or intentions.  Of course it is highly likely that customers (and banks) would get seriously suspicious about this arrangement. Criminals and manipulators could find new business opportunities.

But it is also possible for the customers to arrange it in a clever way and to team up with the help of a trusted operator or an app, for example, and to coordinate their own activities so that the IoB system serves only their purposes. They would gain a major leverage to impact or put pressure on the bank behavior, services or decisions by simply expressing the personal intentions or behaviors. This would open a way to create a kind of an Outside The Wall Street movement. Many a financial, operator and economical policy maker would surely be interested (and worried about it) in this possibility – in good and in bad.

Dangerous and stupid, or both!

Of course IoB is dangerous. “This threatens our privacy”, “It is just another trick to help the big brother watch us”, “You can never know what people really intend to do or what they will actually do”, “This is dangerous”, “It’s a new scale of threat to privacy”, “It is against the human rights” “People don’t want to express their intentions openly” “This will be a paradise for malevolent hackers and manipulators”… Despite these already-heard and imaginary criticisms I believe the IoB is a paradigm worthwhile to consider, to see its potential and to look for the beneficial innovations it could offer – both in theory and in practice. I have already started designing my own toy IoB-experiment.

Nationally I can imagine numerous application contexts, not least in political follow-ups, influencing, analysis, interaction between people and politicians, open government developments, and many more. You might still wonder if this approach can be really different from the already available customer inventories or other means to collect customer data? Yes it can: it is fast, real-time, and relies on acute intentions or acts of people, it scales tremendously, it can be quickly re-oriented, it is easily scalable and it is cheap.

The science of intentions?

There is no science of human intentions. It has remained a poorly charted territory in the psychology while the intentions of politicians, criminals, entertainers and other public figures have remained free material for media speculations and yellow papers. In psychoanalysis intentions have been buried into the mystery world of the subconscious. Modern cognitive psychology as known from the works of Kahneman, for example, have described basic interpretative or choice and decision making behavior and thinking but it does not have much to offer to predict individual, complex behavior like criminal acts, political career choices, and individual or corporate economy behaviors to mention a few. Intentions remain hidden and objects of guessing – unless they can be somehow revealed.

A popular belief among brain scientists and other specialists suggests that it is not objectively possible to know these hidden human forces (see the above link and reference to “unedited behaviors”)  – because we cannot trust what people tell about their intentions and, so the argument goes, we need trustworthy, objective machines or recording systems to reveal them. Surprisingly many seem to trust more in our sweat glands than in our talk.

I want to be clear about this and distance my views from these populist ‘objective’ sciences: Based on my own research experiences with thousands of people in the lab and at their homes or as customers I believe we can trust people when we allow them to be spontaneous, when we use proper methods in interviewing and in analyzing the acquired qualitative data. The present brain sciences are of little help in disclosing complex human intentions in real life and people are often false beliefs about the potential of today’s brain sciences in psychology,  cf.  http://www.innovationjournalism.org/archive/IJO-7-8.pdf  I truly agree with reverend Smith who in his Harlem Church one winter day preached to us “Trust your data! Trust your qualitative data!” He did not know it but he touched on the key discussion topic we had during the journey and when visiting his church.

Individuals can easily describe their intentions before taking actions and better than ever before it is possible to collect this massive real-time and even near-historical intention data. Unfortunately, psychologists are typically uncomfortable with this concept; they prefer the notion of the science of attention, motivation or decision making. But decision making and intentions are psychologically profoundly different: intentions are integrative, internal phenomena that organize the actual decisions but as we so well know (cf. Daniel Kahneman’s or Herbert Simon’s work, for example) they do not directly lead to certain behaviors. Sometimes we do have good intentions but accomplish something totally different. However, I believe that in general there is a fuzzy, but reasonably systematic connection between intentions and the kind behaviors and behavior spaces they precede. Explicit and voiced intentions even have a strong impact on the audiences as we so well know from the politics using threat.

Research on the psychology of life strategies comes close to the science of intentions. Several studies have looked at specific or general problem solving and planning in life (cf. the classic book like Miller, Galanter & Pribram, 1960: Plans and the structure of behavior, or the article by Smith, J., 1996: Planning about life: Toward a social-interactive perspective. Se also B.F. Malle, L. J. Moses & D.A. Baldwin (Eds), 2003: Intentions and Intentionality, MIT Press, for a cognitive and social cognitive analysis of intentional and unintentional behavior. It is my impression – although I’m not a specialist in this field – that these studies do not consider intention very deeply or describe intention knowledge in the framework it is presented here.

Giants will loose their knowledge power?

“Knowledge is power” has become the symbolic statement of social optimism, social revolution, transparency, crowd sourcing, empowerment but also of fear, injustice and threat.  Just look at the Arab Spring, Occupy Wall Street movements, and Wikileaks and it is clear how the dark and bright sides of knowledge mining already prosper together, stronger than ever.

Seeing the avalanche of world knowledge it is astonishing, how few innovations there are for novel data mining paradigms. Visualizations and open data improve fast, massive search, classification, source tagging, and pattern recognition schemes are increasingly used but relevant meanings and significance seem to evade these.  Semantic webs are perhaps the most advanced approaches but even they have problems with context and intention knowledge.

Facebook faces trouble with its mobile advertisement strategies. Google defends its position as the one and only knowledge search and social knowledge sharing emperor. Media world lives under an intensifying turbulence and any speculation in the media about how to best reach the audience is welcomed, entertained and then soon cast away. More intelligent and semantically sensitive systems like Wolfram’s Alpha have not flown yet.

Competing knowledge companies hit the same wall of dynamic world knowledge because the potential pool of meanings and human behaviors underlying any data multiplies faster than the amount of data collected: individuals and social systems have become huge technology-driven, information-generating engines. The more we discover data about the world and use it, the wider will be the gap between what we know and what we would like to know.  Present data mining approaches are loosing the battle.

Intentions and reality

Intentions are the most important guiding data structures in the human mind. For example, when something goes seriously wrong like in the mass murders in Norway, in Finnish Schools, in the U.S. located Sikhi temple, and in the recent killing of the US ambassador in Libya it is natural to look for reasons or psychological predecessors to these murders and to speculate about the intentions behind the acts of violence. Media is lost with these incidents and the journalists conduct post hoc discussions on motivations, political or religious reasons or personal histories of these violent persons. Only rarely they refer to the intentions involved in these acts, as if no such knowledge existed.

Consider the following cases. Firstly, the clerk in the Vatican office who stole secret documents from the papal office is now being charged for the theft. His explanation is that he intended to reveal the misconducts and corruption in Vatican.  But intentions do not seem to matter, even in the papal system, he will be treated as a thief and only that.

The other example comes from Sweden (Sept 2012) where the court in Göteborg freed the suspects (the two men were from Somalia and Irak) who entered a galleria, looking for the artists who had drawn a cartoon of prophet Muhammed.  There was no proof of intended murder although the men had knives with them and they were looking for the artist, who luckily was not there then.  What was their intention? They know it, of course, most people have no difficulties in guessing, and maybe also Muhammed knows it, but the court was helpless, it was not able to deal with he intention knowledge and let them free. Perhaps it was the right thing to do,  I don’t know.

The system of law has problems in dealing with intention knowledge, and in Finland, for example, it is not a crime to prepare a theft or a robber and you can arrange an effective crime exercise in order to fulfill your criminal intention and be even properly prepared for a theft, killing or other crime! Intention knowledge escapes the law authorities totally. (cf., in Finnish about the matter in our Parliament: http://www.eduskunta.fi/triphome/bin/thw.cgi/trip/?${APPL}=utpkk&${BASE}=faktautpkk&${THWIDS}=0.4/1349000344_482448&${TRIPPIFE}=PDF.pdf

Underestimated substance of human knowledge

Without intentions we would be lifeless machines but intention knowledge, the awareness of our own or of others’ intentions, has remained a mystery. Actually, I have never seen anyone include explicit intention structures as components of the human memory and knowledge system. It does not occur in the textbooks of cognitive psychology and when it does it is described under the fuzzy terms of ‘motivation’, ‘attention’ and ‘perception’.  In experimental work it is often something to be controlled away because it can distort the study.

‘Intention knowledge’ is a genuine subjective phenomenon, as familiar an internal aspect of the mind as the feelings of pleasure or pain. Unlike to the experiencing individual, to the data-mining systems, intention has been left outside the focus. Only seldom do they directly deal with the individually verified intention data. And often – like in the law cases – it is seen as too difficult a problem.

In my business, consultation and even research activities I have heard it ad nauseam how it is slow, unreliable or expensive to collect such a detailed and complex data from individuals. Often it is claimed that you cannot trust what people tell about their private mental or physical life. Many demand extremely large samples in order to find anything interesting in consumer behaviors. These assumptions are simply wrong.

It is easy to conduct stupid consumer inventories or other studies, but a cost benefit analysis – which is only seldom conducted – would show how much money is wasted in market research and user studies by not wisely investing in these methods and analysis. Reliability of verbal data can be analyzed and evaluated just like any other type of psychological measurement data. For case examples, see the activities of my research team http://www.poem-research.org/.

For a culturally tuned mind it is no news that intentions matter, only the framing might be different: world literature, art and theatre are built on impressive drama of human intentions, their obstacles, failures, dangers, and other consequences.

The power of intention perception

We perceive intentions. Psychologists might refer to ‘anticipation’ that is an emotion-related phenomenon. However, the two are different phenomena: anticipation describes how a person prepares for any incidence while intention perception refers to the knowledge of the intentions of someone, a person or persons, whose behavior is relevant in the situation. As an example, imagine walking on the street and someone approaches you aggressively, threatens your child or your spouse. You get automatically defensive and tensed, perhaps even aggressive and you want to defend yourself and protect your close ones. What is this ‘perception of a bad intention’? Clearly, it is a peculiar form of human personal knowledge that immediately reserves specific mental structures in your mind for its use and at the same time pushes other ongoing mental and physical processes aside. You get into a state of anticipation.

Similar situations, in a less dramatic form occur frequently in our everyday life when we feel that someone – a person in a street crowd, in a sports event, in theatre play, in an organization or even a leading figure of a nation – reveals intentions to harm somebody. What is common in these instances is the power by which the knowledge – or better, the assumptions – about the intentions touches us automatically and with full force to produce a vigorous reaction.

Clearly, intention perception is more than just recording an event in the world. It is not only opening the gate to the emotional meanings and anticipatory energy but it contains all these psychological phenomena, integrated together in the form of specific type of knowledge. Intention knowledge integrates all our available psychological resources.

Scientists do not know how this significant process happens in the human and animal mind.  Like animals we have to react quickly, not to what we see as the world state, but to perceive what someone is intending to do: a ball thrown at us or an aggressive animal approaching must be immediately perceived. Any deep psychological analysis is time wasted because all this must happen in less than a second as we can see in the fast sports like ice hockey, boxing, karate, squash and tennis. Cf. an example of karate reactions: Had the goal keeper in ice hockey no idea of the intentions of the shooting player he would have no chance to catch the puck coming at the speed of 200 km/hr. It is about half a meter in a millisecond (1/1000^th of a second). The human reaction time at its best is about 100 msec . He must correctly perceive the intention. http://www.budopoint.de/en/science/articles/reaction%20times%20and%20anticipatory%20skills%20of%20karate%20practitioners.pdf

Intuition has become a popular topic in the present management and decision making literature and research (cf. Hodgkinson G. & Starbuck W., (Ed.), (2008) The Oxford handbook of Organisational Decision Making, Oxford University Press, Oxford). One of the reasons to this can be that the basic cognitive psychological knowledge and decision making concepts do not have explanatory power over psychologically complex phenomena like intention and it has led to the use of other concepts like ‘intuition’. In this context it seems to mean that ‘we don’t know what went on in the decision maker’s mind when he made his choice because we cannot look inside him’.

Sometimes the lack of intention knowledge is demonstrated by environmental or technical designs that fail to serve the very basic human need to act according to personal intentions. By overlooking them the designers miss essential aspects of human aims and needs. Our behavior is guided by intentions and we try to fulfill them by looking for or being aware of opportunities to do so.  This is vividly demonstrated by what I call ‘opportunity perception’, our constant habit to look for (to perceive) ways to proceed in life according to our intentions. Bad technical or architectural designs are a temporary  hinder only  (cf. https://gotepoem.wordpress.com/2011/07/12/small-creative-acts-and-opportunity-perception/). Like intention knowledge, its close relative -opportunity perception – is a psychologically integrated phenomenon and there is not generally accepted system model of it.

Truth does not matter (much)?

Watching a movie we know that the subjective feelings associated with ‘intention perception’ are almost the same whether there is objective background to it or not. We all know from personal experiences how strongly the knowledge (which is actually an assumption) of bad intentions moves us. In extreme conditions when life is at stake it gives us the power to kill. No wonder then that painting the picture of  ‘bad intentions’, making people perceive intentions,  is so efficiently used in political rhetoric, yellow paper journalism,  and in various forms of propaganda.

By revealing or communicating imaginary bad intentions the audience is made to react automatically in a way that is more than an intuition because it invites and engages people into internal and often also external action. Fortunately we are also profoundly moved by perceived positive intentions: seeing or experiencing that someone wants to help, love, support, save or just to conduct simple good acts for us.

Making intentions visible

Intention is a psychologically energizing force. Animals have it as many dog owners know; dogs are marvelous in perceiving your intentions – and to react accordingly – and sometimes even better than many humans are.  I remember a story from the book by the Nobel laureate Konrad Lorenz about his parrot that could recognize who would be the next one from the seminar participants to leave the session and it would approach the person even before he/she had announced the intention to leave. The parrot observed the person’s intention.

Because of the predictive value of intention knowledge we have, not unlike Lorenz’s parrot, developed a number of ways to perceive, observe, and recognize the intentions of other people, the opposite sex, our opponents, partners, and competitors. However, intention perception and knowledge processes do not occur at individual level only, it has a significant role at corporate and national level as well. In organizations intention perception (creating and perceiving intentions of the organization itself or recognizing the intentions of other organizations in the environment) is an essential part of the mission-vision-strategy triangle.

A few steps back: knowledge that escapes us

Contrary to the common knowledge optimism, significant knowledge escapes us faster than ever before – not in the absolute sense but relative to what there is and will be for us to know. This paradoxical but creative knowledge explosion is an indirect consequence of the Big Data phenomenon that has taken many by surprise: the increase in the amount of data in the world accelerates faster than the increase in the processing power of all computers that mankind has or will have available. Some factors even amplify the Big Data gap so that Moore’s law is dwarfed by this race, even if it would hold for the next 50 years. Some claim again that it is already breaking http://techland.time.com/2012/05/01/the-collapse-of-moores-law-physicist-says-its-already-happening/. A pervasive challenge will be how to best compute meaningful knowledge from the mountains and clouds of fresh data.

A surprising corollary to the data explosion is that all potential meanings underlying the observable data escape even faster because of the increasingly demanding inverse problems. Apparently similar things and behaviors happen in numbers but because of totally different reasons. If some thinks that by mapping all available data efficiently we will approach the time of revealing all meanings or as it has been said, we live the time of  ‘the end of theory’ (cf. http://www.wired.com/science/discoveries/magazine/16-07/pb_theory/) , they better think twice.

Classic and relativistic worlds of knowledge

Modern ict technology is fast revealing the challenges of knowledge complexity and our concepts of ‘knowledge’ will profoundly change from something to be discovered or constructed  to something that evolves, self-organizes with the help of the knowledge technologies and continuously escapes us by transforms itself. Interestingly, this is nothing new in natural sciences: in quantum physics and cosmology it has been a natural and recognized development.

With all this I mind I was surprised to view the short video http://www.youtube.com/watch?v=mmQl6VGvX-c&feature=share from Google that demonstrated static conceptualizing of knowledge and a straightforward understanding of what is relevant to people and what it could be in future. To use an analogy, they appeared to represent what I see as ‘classic knowledge world’ as opposed to the ‘relativistic knowledge world’ that I promote here.

In the relativistic view the basic assumption is that we live in an expanding knowledge space where the limits of the observable knowledge world become defined by the relationship between the available computational capacity and the speed of knowledge space expansion. To me this appears as an equally complex a problem as the geometrical space-time problem in the general theory of relativity.

Relativistic (but wild) imagination

Why not borrow (and use the analogy for a thought experiment) the concept and role of the maximum speed of light in physics: it prevents us from seeing as far into the universe as we would like to. Similarly, we can imagine a maximum data processing speed and capacity that we have available for accessing the evolving and expanding data universe. Because of this profound speed obstacle we are bound to use our imagination and mathematical models to model and comprehend our universes of stars and knowledge.

But the situation in knowledge processing is not as hopeless as it is in cosmology where we will have serious problems in trying to make direct connection with the distant data sources (far-away-stars, galaxies and universes).  Even though the data sources in our present knowledge world escape fast we can try to minimize the gap between us the escaping wave fronts of knowledge.

Why is knowledge escaping us?

I suggest two main hypotheses concerning the causes of the accelerating knowledge escape:

Firstly, the number of data sources in the world (and in the expanding universe) increases faster than we can improve the available data processing power. This challenge cannot be matched by the best present or near future data processing paradigms (more intelligent and learning data mining tools, massively collective processing arrangements, parallel and distributed computing).  Cf. http://www.weforum.org/reports/big-data-big-impact-new-possibilities-international-development

Secondly, the content quality of the data at the source does not improve fast enough to narrow the Big Data gap by more efficient computing systems and algorithms. This hypothesis is based on the observation of the type of data sources that increase fastest today, due to e.g. dumb use of wide bandwidth, increasing amounts of real-time data, sensor applications and even the coming of the Internet of Things. At present, the standard solutions to the problem of bad quality source data are to increase processing power, to apply better pattern recognition schemes, rely on statistics and learning systems, data triangulation and by collecting suitable contextual data to relax the solution frame.

A delicious example of the nature of this problem is the data processing in the search for the Higgs particle.  It is not possible to tag the data at the source (the particles) and actually (if I understand it correctly) the recording result is not a direct but indirect (subtractive) observation of the assumed particle traces.  But of course, any subtraction result can be produced by an infinite number of observation values – unless there is the one and only correct theory of the world available for the interpretation.

There are other views to the explosion of computational power and Big Data. Some call this “The end of a theory” claiming that more efficient data processing will do it.  There is a hidden snag in this belief, however  (cf. Marissa Mayer from Google, at parc “The Physics of Data” http://www.slideshare.net/PARCInc/innovation-at-google-the-physics-of-data and ….

The most influential technological solutions seem to rely on the idea that we can have access to the increasingly better supplementary data to help solve the inverse problems of the received data. As a practical example from everyday life, imagine a company that wants to serve you optimally and hence, it wants to know your location, what is physically close to you and what is your behavioral history.  What the company actually would like to know is to recognize your intentions and what you need right now and then to use this knowledge in their business as best they can.

Unfortunately, the “human inverse problem” makes this difficult:  you visit the same place for different reasons and in general, different individuals visiting the same place display their behaviors for their own reasons and more often than not it is not possible to deduce these intentions, preferences, likings and decisions from the superficial behavior or history data.

We all know the huge variability of human behavior as a reaction to the same environmental situation. Hence, it is easy to understand why it is so difficult to compute the reasons to these behaviors; it is computationally a several magnitudes more complex a problem than the game of chess.

Why not assign relevant meaning at the data source, as early as possible?

Internet of Behaviors (IoB) amends from the value of intention knowledge and has the power to make intentions visible. The questions I ask here are ”What can we gain from knowing the intentions?” ”What could be done to help recognize relevant intentions?”, and ”How could technology help?”.

The problem of recognizing intentions can be turned upside down and we can think about a simple but valuable solution: to acquire knowledge of the aims, goals and intentions of the people (or of artificial systems, robots) of whose behavior we are interested in. This is not typically done but it can be accomplished by for example, by intelligent tagging of the meaning of the human data and any of its components as early as possible at the source.

The later in the data creation or in the sense-making chain the meaning is introduced to it the more knowledge value drips away from codes. The data becomes irrecoverably biased and the computational problem gets increasingly difficult. It appears quite difficult to imagine how to create self-correcting codes that would recover the lost knowledge of the complex intentions and meanings that can also be dynamic in nature.

The present computation schemes use more processing power for recovering the source meanings that have already been lost from the beginning. Of course, often a typical situation is such that it is not possible to get to the meanings early, but often it is. Perhaps the worst obstacle is that there really are no relevant data structures, or even programming languages (thanks to Paul Suni,  I saw this inspiring talk by Rich Hickey http://www.infoq.com/presentations/Are-We-There-Yet-Rich-Hickey) to express the meaning of sensible human data, especially when it concerns natural behaviors or characteristics (cf. https://gotepoem.wordpress.com/2011/04/05/quantifying-your-self-need-a-data-structure/).

The complexity of the inverse problem of recovering the meaning of the data at its source is increased by the computations applied to it after its coding at the source.

It is a most ambitious computational problem to try to uncover the intention of a behaving system, be it natural or artificial in nature.

The walking pilgrim with an intention: We would like to know why a pilgrim walking his Camino chooses his own specific trail.  Because it is an open problem it is not a general case of the traveling sales man problem: we want to know why he has produced the specific movement patterns on his pilgrimage.

Typically the pilgrims on the famous Camino de Santiago all have the same goal: to reach the Cathedral de Santiago in Northern Spain. They take a long walk that can be anything from 3000 km to 50 km long. We know that the motivations behind the nearly identical movement patterns can be totally different.  Interestingly, when starting their journey they pilgrims are asked by the church clerks: “What is your motivation, religious, sports, spiritual?”

To solve the problem we could use an accurate gps system and map the pilgrim’s  location on the Camino at every point of time to obtain data that accurately describes each individual route (x, y, t). From the data recorder’s point of view, a dense sampling and rich contextual, supporting knowledge should help to reveal the motivations underlying his walking behavior. Knowing a number of constraints, like the knowledge that the pilgrim is a tourist, a priest, a sportsperson, what time of the day it is, is he hungry, or does he live near by relaxes the problem and allows the exclusion of a number of false interpretations.  In some cases, and in the average data we might arrive at reasonably useful interpretations of the walking pilgrim with an intention on his Camino. But in general this is not so.

From the pilgrim’s point of view, the walk has a number of intrinsic, evolving and fulfilling motivations. Like any healthy person, a pilgrim or not, it is highly likely that his mental and spiritual drivers are dynamic and change continuously (cf. https://gotepoem.wordpress.com/2011/10/13/entropy-of-the-human-soul). Every time his journey motivations change, the walk data analysis becomes irrecoverably problematic: an intentional pilgrim walking does not remain a stationary system. After all, for most pilgrims this is the spiritual idea behind the walk. It is even risky to consider him physically a stationary system and certainly it is not reasonable to make this assumption about his mental states. As a consequence, it becomes practically impossible to recover the original meaning of the (x,y,t) data without accurate and dynamic pre-knowledge of the pilgrim and his evolving intentions.

There is no computational method to solve the problem of the intentional pilgrim– unless we have relevant data available at the source.  Realizing this made me think about “The Internet of Behaviors”, (IoB) and to ask the question: What if we could have valuable meaning data at the source? How and when could we record it and how could we use it? Now after some time since the thought experiment I’m convinced about the computational potential of IoB.

In the traditional mapping of the pilgrim’s trail we loose knowledge content each time a computation is performed to calculate the distances, velocities, accelerations, turnings or other aspects of the behavior pattern. For example, when two dots on a map are computationally connected there is a high probability that this most simple computation cannot be mapped at all to the internal space of the pilgrim walking. In other words, the relationship between the computed movement patterns and the internal states of the pilgrim remains fuzzy.

We can use clever movement pattern characteristics like entropy, for example, to model the movement types (cf. Särkelä et al., International Journal of Human-Computer Studies, 2009) but every computation introduces noise to the meanings of these movements at their source. Hence, it becomes increasingly unclear why these movements patterns have occurred.

Preserving the significant meaning at the data source may not be not a mainstream idea (I do not know what exactly has been accomplished in this field), and I’ve been repeatedly puzzled why this view has received so little interest even with the presence of clever sensing systems and other data collection innovations.  It is almost as if the computing systems were designed to accomplish just the opposite: most of the computing power is devoted to clean up data because of its bad quality at the source.  Mathematically this can be elegant (to solve difficult problems) and sometimes it is just a necessity of life because we do not know more about the data itself and its creation. But what if we could know more? What if we could know what is meaningful data – very early when data is generated?

There are well-known technical and theoretical solutions to this: Real-time operating systems declare significant events so that when an important external or internal event takes place the system program reacts to it according to the nature of this interrupt or ‘flag’. However, within the human and social sciences this kind of thinking is not mainstream and it has not been used to model human behavior and its control.

Apps?

Some novel apps indicate that innovators are realizing the benefits of improving the quality of the data at the source.  See for example, http://www.thinglink.com/ that allows tagging (albeit later in the process) of images so that they could provide relevant content information to the viewers. This approach can be improved by suitable means. In the following thought experiments I’ve described some direct ways to achieve similar advantages.

To reiterate the idea of Internet of Behaviors:

IoB aims at tagging the essential meanings of a specific behavior pattern at its source and as early as possible  (related to the behavior of a human, machine, organization or data processing chain, for example). By doing this IoB opens a gateway to that specific behavior pattern to be used for beneficial purposes.

IoB is a communication scheme according to which a large number (e.g. from a few to several thousands, see e.g. IPv6) of IP or ‘IoB’ addresses can be reserved by individuals, communities, couples, teams, organizations, or interested operators who use these addresses to refer to the specific behavior patterns that they want to declare explicit and open for a trusted or public use.

IoB is analogous to the Internet of Things, with the difference that IoB refers to behaviors that have been pre-defined, defined on-the-go, or otherwise specifically declared by someone. In other words, it carries human knowledge of what is the meaning and significance of the addressed behavior pattern. Unlike in the Internet of Things the addressed behaviors need not be exact in nature or fixed. IoB relies on the intelligence and knowledge of the operators (individuals, social entities, organizations…) to denote meaningful behavior patterns, either actively or by a suitable technical scheme. We all know the social value of such knowledge.

However, the owners of the addresses (individuals, communities, designers, engineers, artists, journalists) know, believe or expect that it is somehow beneficial to identify and express their specific behaviors for which they have reserved the addresses. When these behaviors occur the IoB opens access to the data and processes relevant at the occurrence of that specific behavior.

IoB scheme is dynamic in the sense that unless the specific behavior occurs, the address is not available. It is a pattern recognition scheme with the specific property that the recognition is based on simple (voluntary) indication by one or more individuals, by a sensor system, a recognition sw, or by any imaginable human or technical means.

What could be achieved with the IoB scheme?  A lot.

A few days ago I was listening to a podcast interview (from Carnegie Council http://www.carnegiecouncil.org/resources/multimedia/20120716/index.html of Mark Nelson and Margarita Quihuis from Stanford Peace Innovation Lab http://www.stanford.edu/group/captology/cgi-bin/peaceinnovation/?page_id=35 . Margarita mentioned the facebook experiment that they were involved with where it is possible to follow ‘likes’ over conflict boundaries, for example how many people are showing this behavior at any moment (cf. http://peace.facebook.com/?ref=ts). But this is just the beginning: what if we could join the specific behavioral situations where these positive engagements occur? What could we then accomplish?

If we know enough of any specific behavioral situation, be it a human or a machine, it is possible to compute relevant knowledge from this data. With IoB there are benefits:

1. When there is not enough contextual data to deduce the meaning of behavior B_n of an actor in a specific situation (“I’m in the library preparing the manuscript X”), then the indication of the actor state S by IoB_S  that the actor itself (ideally) provides helps achieving correct information about B_n without complex computations.

2. The intentions of an actor are difficult to deduce by any computational means in natural life situations. IoB can be an indication of these intentions and help effective computing of the meaning of the observed behavior.

3. Knowing the intention of an actor provides a valuable means for proactive or even predictive analysis of the future events and behaviors. When there is a large number of actors this becomes especially valuable.

4. People are not very good at tuning and managing a number of applications to monitor relevant behaviors of other people or devices. But they are very knowledgeable even to the smallest detail in indicating their own behaviors, and even their own thoughts or feelings if allowed to do that in a simple and secure way.

Example apps that could fly

Let’s assume that a child has a mobile system attached to his school ubi-bag. Every time the bag moves from his home or from the school it opens an IoB address (only) to the parents who now have an IoB address to the school trip. How they want to use this knowledge is a matter of habits, needs, ethics, and imagination.

A journalist working in the middle of a conflict area can indicate his or her intentions to collect material of certain type of incidences. By opening his IoB he can get real-time support, materials and even protection and guidance from trusted parties like other journalists.

Imagine that every time to you read a specific issue of a magazine you could indicate it with your mp or the magazine itself would know that you have it in your hands.  The publisher would then know that you are “engaged with magazine X” or even in more detail what is the article you are reading or the specific material in the magazine that you are searching for. The IoB address that you have declared for this behavior (or the publisher has arranged it in a clever and even commercially viable way) lets the publisher offer any service to you that is related to this reading, be it any journalistic content or adverts. Only imagination is the limit to the potential and commercial ways to arrange for this new type of mediaspace to be shared with the publisher.

What if everyone coming to an airport would open his information about the travel intentions and on-going activities to the travel offices? By knowing the intentions and on-going travel activities of their customers the airports could be better prepared – in advance – to use their limited resources, help them in receiving special guests or people with problems and simply to improve their overall service relevance.  The real gain would be the advanced knowledge, an expected-behavior map, that is, a better recognition of the customer data at its source. There is no magic knowing the check-in gates but the ways and intentions that the travelers have in getting to them can carry valuable and even totally novel knowledge for the service providers – and to the travelers themselves. Present net-services are rather clumsy.

Imagine that you are studying how in your organization innovations are created. It is difficult to predict exactly where and how this happens, but before the study, on the basis of interview data it is possible to guess, which people, places, circumstances and behaviors are likely to produce that kind of activities. Based on that knowledge, it is possible to span a set of IoB addresses pro-actively and then collect data (location and other activity records, identifying other people involved, launching a questionnaire) only when the persons studied have indicated that such relevant episodes have happened or are about to happen.

Imagine that activists share the address of their next operation, like participation in a demonstration. When everyone participating opens the IoB gate to this occasion we will have thousands of eyes, ears and recording devices to know what is really going on there. Address & code traces will be left of these activities as well so that if anyone would want it, a shared coding system could help integrate vast amounts of eye-witness and other data. Naturally, the ID protection is a most important requirement.

Imagine that it would be possible to divide any problem into sub-problems (with IoB codes), just like it is done in Microtask (http://www.microtask.com/). Then, imagine further that you had access to anyone who, at any moment, happens to be engaged with solving that specific part of the problem. Imagine even further, that you would have direct access to the solutions that these people are producing or have produced. What could we accomplish with such a system? A lot, especially when we build a system according to which large problems can be compartmentalized.

Finally about band-fan management: what if your favorite band had a gateway to offer to those attending their performances? There are reasons why they might want to open it, in whatever way they see useful, also to audiences (and record markets on the other side of the globe) outside the concert halls?

… and then there is an unimaginable amount of app potential for cars, games, new media, research, and mobile phones, of course. Somewhere IoB will start living.

Edit on 31st October 2021: I replaced the ‘IB’ with ‘IoB’ to make it compatible with current practices. Contents have not been edited.