Reciprocity compels us to treat everyone as an equal, the proverbial "Golden Rule".  Anticipation is the system of rules, knowledge, clues and taboos that enables us to predict how a human being will react in a certain situation, and ultimately how to evaluate and understand his or her actions in light of said rules.  The normative part of ethics if you will.  This simplification is a model of sorts to undestand Ethics as a whole.  The minute aspects of all ethical systems are varied.  But probably the majority of them will lead to this Ueber-model. 

The Model as such is a tool that enables us to understand complex behaviours when there is a lack of data or of time to analyze every aspect of a problem.  A model, if designed properly, can account for a large number of individual cases, but is only an approximation of the truth.  Many cases will be left out or details will not be considered.  And it is in the details that we really obtain meaning and identity.

So it is with the principle of Anticipation.  We create a model of the human being in a certain ethical, cultural and social setting.  We then use this model to predict and judge behaviours of individuals and groups.  The catch-phrases of the anticipatory model are everywhere in our culture as memes, sayings and "Binsenweisheiten".  But it goes very much beyond phrases and memes.  The anticipatory system is the basis of the literary canon of western civilization, from Homer, Aquinas and Isidore to Freud, Shakespeare and Cervantes.  The Canon is our working model and hypothesis of the (western) human being.  It explains the passions, dreams, behaviours, love, comic... everything possibly human.  And it does so by force of repetition and approximation, rather than by facts and experimentation.   One of the books of the canonic scholar and literary critic Harold Bloom is aptly titled: Shakespeare: The Invention of the Human.

But now fast forward.  Chris Anderson published recently an article in Wired questioning the validity of scientific models.  His point is, why do we need models if we can go for the real thing?  Today we have enormous amounts of data and information available, and moreover, have the computational capability to process an interpret such information.  This is going to have an impact on how we see and comprehend the world.

And what happens to the human model, the canon and the ethics?.  We have not only information about wordly things like traffic and weather, we also have a lot of information about how humans behave.  What they write and care about, what they think of events, how they react to catastrophes and how they fight against political systems.  Even information on how they buy, whom do they know and speak to, and where do they travel to.  The life of the "nomadic Cyborg" (W.Mitchell) is ever more lived out in the electronic Landscapes, Opens and Commons.

The practice of extracting this information, of making sense of all the data in the computer systems is called "Reality Mining":  In other words, how to obtain a picture of reality from raw data in a non-normative way.  Don't make assumptions (i.e. models) about reality, just extract reality from the data and understand it.

Alex Pentland of MIT Media Lab calls this "Honest Signals".  It is about how we can extract a clear picture of the human being, its environment and its social operations.  With this information we can then make true statements about the affairs of the person and the social group.  This could have applications in a wide variety of situations, like to determine the social sentiment of a group (a happy group or a civil unrest) from the tone of their voices.  Or the emergence of public health problems and epidemics to be able to respond faster.

Nevertheless the demise of the model of the canon and the rise of Reality Mining in its stead opens up a series of questions, ethical, political and legal, that must be pointed out. 

First of all: Privacy.  Do we as citizens want to be exposed?  One of the rights of citizenship is the one that creates my private sphere.  And consequently the social civic sphere. 

Second: The political form of free will.  This area where we as a society still have the option to act out of free will and not in a perfectly anticipated way.  What happens to free will if there is no anticipation, no measure of freedom to act outside of predictability?  We need some degree of confidentiality to mantain the anticipatory system of Ethics.

Third: Social Capital.  This is an important question for the crafting of a digital citizenship.  The right of ownership and privacy of our personal information.   Who owns my social links and spheres of influence?  Everybody uses Social Capital for personal advantage: to obtain a job, to do sell or buy stuff, to be part of a Commons like the Internet.  If my Social Capital is in the hands of social networking sites, should I receive compensation?

Here are two interesting quotes.  One from Dr. Pentland speaking of Reality Mining, of which he says that "it's an interesting Gods-eye view" (MIT Technology Review, 2008, TR10).  Of course he is referring to the possibilities to do good.

The second from Bjarne Stroust-Rup, when asked by MIT Technology Review (Jul/Aug 2008) about the Future of the Web: "The total end of privacy.  Governments, politicians, criminals, and friends will trawl through years of accumulated data (ours and what others collected) with unbelievably sophisticated tools.  Obscurity and time passed will no longer be covers."

Obscurity and Doubt are a necessary part of every ethical and political system.  It is a precondition for freedom and justice.  We need to make sure that we don't lose all of it following the New Enlightment movement of Reality Mining.

[slideshare id=509642&doc=05intelligentsensornetworksmanuelangelpatricio-1215802352026769-8&w=425]

Estos sistemas permiten automatizar muchas tareas onerosas en sistemas de vigilancia, por ejemplo, y cambiar completamente las funciones de los recursos humanos, que tendrán a su disposición información más elaborada. También tienen aplicaciones en sistemas medioambientales.

El paradigma de sistema de inteligencia ambiental se puede ver como un sistema compuesto por redes de sensores inteligentes (i.e con inteligencia embebida) para capturar la información de contexto de los grupos de usuarios y usuarios individuales, incluyendo como de capital importancia los redes de sensores visuales. Otro componente es un sistema de fusión de sensores y construcción de información de contexto, un ecosistema de interfaces multimodales inteligentes y una red de agentes distribuidos a varios niveles para gestionar, adaptar y realizar las tareas encomendadas al sistema. Este paradigma se puede aplicar a sistemas domóticos, sistemas de vigilancia o sistemas de control medioambiental.

Es interesante ver una aplicación más sofisticada de las redes de agentes inteligentes que hemos visto en el taller del curso. Hemos pasado de ver aplicaciones de juguete como la del mercado, la de la sabana, la del tute, a la gestión inteligente de cámaras inteligentes con servicios como la detección y seguimiento de objetos (personas, vehículos, etc.) capaces de colaborar y comunicarse entre sí para realizar esas tareas.

Se mostraron ejemplos de desarrollos realizados por la Aware-Home Research Initiative. La filosofía que se persigue ahora es monitorizar el comportamiento de colectivos (en el caso de los hogares: las familias) para adaptar mejor el interfaz de usuario de los sistemas domóticos tratados como sistemas de inteligencia ambiental.

Un concepto importante que subyace en estos sistemas es el de inteligencia perceptual acuñado por el profesor Alex Pentland del MIT.

El problema de escalabilidad de las redes de sensores

públicos como calles, plazas, estaciones de tren, autobús, metro, etc. Esto es debido a que estos sensores se han abaratado considerablemente. Se puso como ejemplo paradigmático la ciudad de Londres, con más de un millón de cámaras. Surge el problema de cómo gestionar esa gran cantidad de cámaras y cómo analizar toda la información recibida de esa miriada de sensores visuales en un sistema de monitorización. El uso tradicional de seres humanos para la tarea de interpretación de las imágenes se ha convertido en una cosa imposible no ya por el coste en recursos humanos sino por la capacidad limitada en tiempo de la atención de un ser humano. La solución pasa por automatizar la obtención y el preprocesado de las imágenes y la presentación al usuario humano de una información más elaborada y simplificada.

NOTA: Según he podido ver en la web del profesor pentland, a esto se le llama ahora Reality Mining.

Aplicaciones Reales en el Mundo Real

A día de hoy y a pesar de la publicidad ofrecida por las compañías que se dedican al negocio de los sistemas de monitorización (se mencionó a una empresa famosa surgida del mundo académico como ObjectVideo) no se ha conseguido ni mucho menos que los sistemas funcionen en entornos complejos. Esto se debe a que las condiciones de laboratorio son condiciones controladas donde es fácil obtener la información de contexto precisa que necesitan estos sistemas para funcionar de una manera fiable. No obstante los usuarios han encontrado aplicaciones alternativas a estos sistemas imprecisos y limitados.

Como anécdota se contó la de un grupo musical alternativo que se dedicó a grabar un vídeo musical promocional actuando en varios escenarios delante de cámaras de seguridad y aprovechando la ley que exige a los usuarios a los que se graba la posibilidad de recuperar las imágenes donde aparece dudieron montar el vídeo.

Otro ejemplo es el análisis del comportamiento de colectivos de objetos como personas o vehículos. Esto puede ser aprovechado para la gestión de infraestructuras como carreteras, metro, etc. o en centros comerciales, donde se puede determinar las zonas de mayor tráfico de personas para labores de merchandising. Otro ejemplo más espectacular es el análisis de actividades deportivas, el seguimiento de jugadores de fútbol o baloncesto o el comportamiento colectivo de los equipos en los partidos. La información obtenida de estos sistemas puede servir para dar noticias curiosas como el número de kilómetros "recorridos" por el portero de la selección española en la eurocopa o para obtener información de jugadores concretos con valor para entrenadores y cazatalentos.

Evolución de los sistemas de monitorización de vídeo

El ponente explicó la evolución de los sistemas de monitorización desde los vetustos circuitos cerrados de televisión analógicos (CCTV) hasta los actuales completamente digitales como cámaras inteligentes (con agentes inteligentes empotrados con capacidades de visión artificial) conectadas por redes inalámbricas e IP a centros de proceso de datos donde se analizan las imágenes y se elabora el conocimiento e información de contexto necesaria para la aplicación en cuestión.

Sistemas MultiAgente para Redes de Sensores Visuales

En algunas de las transparencias se menciona el trabajo del ponente en este campo en donde abogan por los sistemas multiagente (MAS) como solución al problema de la gestión de sensores virtuales, alineación temporal y espacial, y coordinación en tareas de detección, seguimiento, localización, etc:

Multi-agent framework in visual sensor networks
EURASIP Journal on Applied Signal Processing archive
Volume 2007 , Issue 1 (January 2007) table of contents
Pages: 226 - 226
Year of Publication: 2007
ISSN:1110-8657

NOTA: He buscado en la red sobre el tema del paper con muy interesantes resultados.

Otro objetivo, que es aumentar la capacidad interpretativa de los sistemas, se consigue aumentando la variedad de sensores utilizados (y su adecuada fusión), la representación multimodal de la información obtenida y la construcción de una información de contexto de calidad. El ponente indicó que la incorporación de esta información de contexto va a suponer una revolución en estos sistemas. La interpretación en alto nivel de las escenas va a requerir de esta información de contexto (¿usando CBRs u otras técnicas quizás?)

Trabajos desarrollados en el GIAA

El ponente comentó algunos trabajos que demuestran las mejoras que se pueden obtener utilizando redes de sensores visuales con agentes inteligentes autónomos embebidos. Uno de ello consistió en construir un sistema de tres cámaras para la identificación de individuos a través el iris. Se consiguió un sistema con tres cámaras distribuidas por un laboratorio que eran capaces de localizar y obtener una imagen del iris de una persona a 7 metros de distancia sin necesidad de que la persona mirara a ninguna de las cámaras. Las cámaras colaboran entre sí para localizar al individuo, su cara y sus ojos y obtener una imagen de calidad el iris de uno de sus ojos. La distancia normal de obtención de la imagen del iris es de apenas un metro, mirando sin moverse fijamente a la cámara.

Técnicas Avanzadas

Una técnica de seguimiento muy interesante es la de filtro de partículas que permite definir para cada objeto a seguir una serie de puntos sobre los que se estima sus características cinéticas. Esta técnica ha permitido obtener resultados de detección y seguimiento de múltiples objetos simultáneamente en escenas complejas. El problema que tiene es su coste computacional que ha permitido analizar solamente del orden de 2 fotogramas por segundo. Esto significa que los resultados que se mostraron en los clips de vídeo que nos mostró el ponente en el que se veían imágenes con vehículos y personas en movimiento que eran localizados, marcados y seguidos, fueron el resultado de un costoso análisis offline de las imágenes. Es necesario procesar al menos 7 para obtener un resultado aceptable en imágenes de vídeo en tiempo real.

Hay otras técnicas matemáticas estadísticas de obtención de características de imágenes de vídeo, especialmente caras, como la de Viola-Jones.

Agentes Intencionales (Agentes deliverativos BDI)

Las cámaras inteligentes tienen embebida la capacidad de detectar, reconocer y seguir objetos y personas. Esa capacidad está en forma de servicios que implementan agentes inteligentes autónomos deliberativos (BDI) El ponente mencionó el trabajo de Michael Bratman, un eminente filósofo, sobre la base filosófica sobre la que se sustentan estos agentes deliberativos:

Intention, Plans, and Practical Reasoning (3/15/99)

Michael E. Bratman

Bratman develops a planning theory of intention. Intentions are treated as elements of partial plans of action. These plans play basic roles in practical reasoning, roles that support the organization of our activities over time and socially. Bratman explores the impact of this approach on a wide range of issues, including the relation between intention and intentional action, and the distinction between intended and expected effects of what one intends.

ISBN (Paperback): 1575861925

Subject: Philosophy; Intentionalism; Planning

Yo he encontrado otra referencia interesante que vale la pena leer:

BDI Agents: From Theory to Practice
Anand S. Rao and Michael P. George
Australian Articial Intelligence Institute

Demostraciones

El ponente puso unas cuantas demos en vídeo de sistemas de detección, reconocimiento y seguimiento.

- Los cascos de realidad aumentada de los bomberos
- Los vídeos demostradores de ObjectVideo
- ...

This is the core of online marketing's advantage over traditional media. But what is happening where most of us still live? i.e the physical world.

Well, we have various profiling tools, such as affinity cards. So we can figure out how people are spending. But how did we land up in the store that day? A coupon code perhaps? We cant really be sure. We know things like 12m people are supposed to have visited this location.  But then we don't really know where they went or where they came from (we think perhaps 25% are international based on surveys, for example).

But this is starting to change. Mutopo is working with one client to track in store movement, using signals from cell phones. The approach lets us show a location owner how people move around the space, not unlike what I might see from online analytics - navigation paths, entrance and exit points, time spent at certain locations etc. And so now, I can make changes in reality and measure the responses, not just changes in sales.

While we are working on a very local scale, we were excited to learn about Prof Tony Jebara's new project, Sense Networks. Tony and his team are harvesting a variety of data sets to understand what people are doing - they are, in effect bringing online analytics into the real world.

Things like the most searched items on Google Trends might have realwold analogs such as the most visited restaurants. Or conversion reports might now be possible from outdoor campaigns, as you can get a real sense of the number of people who might have walked past a specific location.

The image on the left shows an example application to show the "hotspots" in San Francisco. These are literally the places you want to be if you are asking the question: where is everyone going tonight? Yes, its realtime. You can learn more about this app at Citysense.

It might be possible to know for sure how many more people in New York City have chosen to bike to work. I can start to see if more people are going to the new Ikea in Brooklyn instead of alternatives in New York City, such as Bed Bath and Beyond or the Container Store. Where I might have used Google Analytics to benchmark my site, now I can do the same for my store.

And we can now play what-if, in the real world. What if we:

• change the layout of the store?
• place new promotional signage at the hallway?
• invest in signage alongside the highway?
• purchase the locaton on 26th and 5th?
• notice that more people are starting to cycle in NYC?
• see more people dining out in a new area of town?
• see more people going to Trader Joe's than Wholefoods?

Its feels like we are on the verge of a significant change in how data from cell phones, GPS devices and the like can be analyzed in new and interesting ways. Good luck to Tony and the team at Sense Networks. We cant wait to see what people are going to do with your analysis.

A New York based company called Sense Networks recently launched two new services. One is an platform called Macrosense that (anonymously) tracks where people move around via their GPS or GPS enabled phones, WiFi, triangulation, etc. The the other is a mobile application called Citysense that enables people who opt-in to being tracked to access data about what's "cool" based on where people are going - and how often.

While the latter application falls squarely within the newly-hot geotagging movement and the now-maturing social networking spaces, it exists mostly to entice people to opt-in to the service - if you allow Sense Networks access to your data, you get access to Citysense. This model apparently works better than paying people outright for their data.

However, we think the data mining platform is more intriguing. The company lists many interesting applications for it such as identifying "influence points" along popular routes or qualifying demand (and the elasticity of demand) for places, activities and services by income level. The platform might also be useful in qualifying respondents for consumer insights work. Instead of self-reporting, one could use this system to create a recruitment center based on people opting-in to allow you access to their movement data so you would know for certain the potential respondents had actually been to Acme Retail Store before recruiting them for an IDI or focus group.

Sense Networks insists they aggregate the data anonymously (and one of its founders now consults for them exclusively on privacy issues). Still, at some point in the process, the system must collect unique data such as income, gender or age before it scrubs and aggregates this data. As we have mentioned before, we think that the advantages of persistent identify and the ubiquity of surveillance will continue to trump privacy issues. Still, opting-in to a system that tracks your every movement would seem to be a hard sell. It will be interesting to see how consumers respond to Sense Network's inducements. If consumers derive enough value from trading on their geodata, you can be sure the big Telcos will follow Sense Network's lead.

While the latter application falls squarely within the newly-hot geotagging movement and the now-maturing social networking spaces, it exists mostly to entice people to opt-in to the service - if you allow Sense Networks access to your data, you get access to Citysense.  This model apparently works better than paying people outright for their data.

However, we think the data mining platform is more intriguing.  The company lists many interesting applications for it such as identifying "influence points" along popular routes or qualifying demand (and the elasticity of demand) for places, activities and services by income level.  The platform might also be useful in qualifying respondents for consumer insights work.  Instead of self-reporting, one could use this system to create a recruitment center based on people opting-in to allow you access to their movement data so you would know for certain the potential respondents had actually been to Acme Retail Store before recruiting them for an IDI or focus group.

Sense Networks insists they aggregate the data anonymously (and one of its founders now consults for them exclusively on privacy issues).  Still, at some point in the process, the system must collect unique data such as income, gender or age before it scrubs and aggregates this data.  As we have mentioned before, we think that the advantages of persistent identify and the ubiquity of surveillance will continue to trump privacy issues.  Still, opting-in to a system that tracks your every movement would seem to be a hard sell.  It will be interesting to see how consumers respond to Sense Network's inducements.  If consumers derive enough value from trading on their geodata, you can be sure the big Telcos will follow Sense Network's lead.

This quote appeared in that post:

The days of standalone GPS devices may be numbered. At least, that’s what cell phone vendors and service providers would have you believe.It turns out they may be right. In the span of about a year, we went from almost no GPS-enabled handsets to close to half of all available models including the navigation features. And that number will only increase, with GPS radios becoming standard in 2008 much like cell phone cameras became the norm back in 2004.

Today, the New York Times reports on the development of applications that are taking advantage of the ubiquity of GPS devices.

Via this story, we see that it hasn't taken long for the engines of business to find ways to capitalize on the technology. A few tastes:

We’re in the midst of a boom in devices that show where people are at any point in time...

Such data could redefine what we know about consumer behavior, giving businesses early insight into economic trends, better ways to determine sites for offices and retail stores, and more effective ways to advertise.

Just this month, the journal Nature published a paper that looked at cellphone data from 100,000 people in an unnamed European country over six months and found that most follow very predictable routines. Knowing those routines means that you can set probabilities for them, and track how they change...

It’s hard to make sense of such data, but Sense Networks, a software analytics company in New York, earlier this month released Macrosense, a tool that aims to do just that. Macrosense applies complex statistical algorithms to sift through the growing heaps of data about location and to make predictions or recommendations on various questions — where a company should put its next store, for example...

Mr. Jebara, who is also an associate professor of computer science at Columbia University, says the key to drawing such conclusions starts with having very large sets of data that go back several years. Sense’s models were developed initially from sources like taxicab companies that let it look at location data over such a period. Sense also uses publicly available data, like weather information, and other nonpublic sources that it would not disclose. “We had three-quarters of a billion data points from just one city,” Mr. Skibiski says.

Mr. Jebara’s statistical models interpret those patterns and look at whether they correlate with things in the real world, like tourism levels or retail sales. The algorithms are complex. Even so, the model doesn’t work for everything Sense tries it on, often because more data is needed. But Mr. Jebara says that when it has the data, the model works well. Several hedge funds made an investment in Sense earlier this year.

The Macrosense tool lets companies engage in “reality mining,” a phrase coined by Sandy Pentland, an M.I.T. researcher who was also a co-founder of Sense and now advises it on privacy issues...

“The reality is that location data is new, and we don’t have 10 years of history to work from,” says Ted Morgan, the chief executive and founder of Skyhook Wireless, which sells a service that lets people use WiFi network access points to get information about their location.

“But if their algorithms can do the things they say, we’d probably do a lot with them,” Mr. Morgan says.

Read the whole thing.

Friends, things are about to get very, very interesting.

Cross-posted at Gates of Academe.

In the past, cellular companies who own this information have not used it for fear of alientaing subscribers and also due to privacy infringements. As the viability and marketing potential of the information increases, cellular companies are begining to collaborate with business partners to market their respective products based on location and calling habits for just-in-time marketing. There are many other just-in-time applications for this information and the wireless medium and it is only a matter of time before it becomes another "viral" advertising medium!

The potential for "Reality" mining is only begining .....

Enzimas celuloticas
Mineração Real
Connectomica
Aplicações web offline
Transistores baseados em Graphene
Magnetometros atomicos
Transmissao de energia por ar
Nanoradio
Chips probabilisticos
Modelagem de surpresa

Mais detalhes aqui

This year's list includes a technology for wireless power; there are quite a number of these developments around these days, including some the TR10 missed such as Splashpower in the UK. It includes "Modelling Surprise" – not magic, but developing a scenario modelling methodology that takes into account disruptive surprises of the past. Also modelling uncertainty is probabilistic chip design, which recognises the range of computational areas where absolute precision is un-necessary and thus enables power consumption of devices to be reduced: this technology may also come into its own as component sizes continue to reduce and the physics of uncertainty come into play in mainstream design.

Sandy Pentland's Reality Mining is also in the list: by enabling mobile devices to "recognise" each other when they're in range, data can be gathered about the social or professional encounters between their owners. Knowledge of their personal networks can then help facilitate serendipitous meetings (think "I didn't know you'd be here!")

Visit TR to review the complete list.

Links:

• Technology Review's Ten Emerging Technologies of 2008 (Tech Review, March/April 2008)

• Splashpower

Every time you use your cell phone, you leave behind a few bits of information. The phone pings the nearest cell-phone towers, revealing its location. Your service provider records the duration of your call and the number dialed.

Some people are nervous about trailing digital bread crumbs behind them. Sandy ­Pentland, however, revels in it. In fact, the MIT professor of media arts and sciences would like to see phones collect even more information about their users, recording everything from their physical activity to their conversational cadences. With the aid of some algorithms, he posits, that information could help us identify things to do or new people to meet. It could also make devices easier to use--for instance, by automatically determining security settings. More significant, cell-phone data could shed light on workplace dynamics and on the well-being of communities. It could even help project the course of disease outbreaks and provide clues about individuals' health. Pentland, who has been sifting data gleaned from mobile devices for a decade, calls the practice "reality mining."

Reality mining, he says, "is all about paying attention to patterns in life and using that information to help [with] things like setting privacy patterns, sharing things with people, notifying people--basically, to help you live your life."

Researchers have been mining data from the physical world for years, says Alex Kass, a researcher who leads reality-mining projects at Accenture, a consulting and technology services firm. Sensors in manufacturing plants tell operators when equipment is faulty, and cameras on highways monitor traffic flow. But now, he says, "­reality mining is getting personal."

Within the next few years, Pentland predicts, reality mining will become more common, thanks in part to the proliferation and increasing sophistication of cell phones. Many handheld devices now have the processing power of low-end desktop computers, and they can also collect more varied data, thanks to devices such as GPS chips that track location. And researchers such as Pentland are getting better at making sense of all that information.

Broke and backwards as I am, I actually don't have a mobile phone (or cable, or a super-high-speed Internet connection, or ...). But as the article notes, there are lots of issues here about personal privacy: if, for example, it's really possible - as in the work of one of the scientists that the article discusses - to "use cell-phone data to improve existing computational models of how diseases like SARS spread" by providing "detailed data on where and with whom people spend their time", who is it that is going to have this information, on what grounds are they going to obtain it, and exactly what is going to be done with it once it's been "mined"? It seems to me that the last thing we need is a bunch of super-smart scientists, whether in government or in private industry, keeping track of whom we're talking to in order to track the causal chains involved in such things as the spread of disease.

What's so unsettling about this technology is that since it doesn't involve anything as blatantly invasive as wiretaps and - so far - isn't being used for especially scary-sounding purposes (though using speech analysis software to look for signs of depression, or monitoring a phone's motion sensors for changes in gait that might be suggestive of Parkinson's, is the kind of thing that makes me shudder), most people aren't going to be inclined to worry about it very much. But these are serious issues and promising technologies, and so - as Pentland says at the end of the article - it's imperative that we all start having a conversation about how we're going to regulate it. Manzi's proposal that we create a "market" for personal information, allowing people to charge commercial entities to access their personal information and plug it into their models, is a good one, though will it work when the reality-mining entity in question is a branch of the federal government or a do-gooding non-profit? What's going to happen when it inevitably turns out that reality mining can be used to predict terrorist attacks, or to track and treat the outbreak of serious diseases? Will the fact that "merely" analyzing your overall patterns of conversation and motion can give rise to such an accurate and powerfully predictive picture of your behavior make the need to actually listen to your conversations (and so to go through the red tape required to do that) seem a bit primitive? And is engaging in such "content-free" analysis somehow less of an invasion of one's privacy than actually listening in, even if the understanding of one's (past, present, and future) behavior that it enables is even more robust than that provided by wire(less)-tapping? Finally, if it turns out, as it surely will, that reality mining can only really when information about a sufficiently large segment of the population is plugged in, what's going to be done about those of us who don't want our reality to be, well, mined?

Difficult questions, these. All I know is that if I ever do cave in and buy a mobile phone, it's going to be a Stone Age model like this one.