Post-humans Sighted

I’ve exorcised my demons and put them between covers under the title Silicon Simulacra: Post-humans of the Machines Worlds.   If you don’t want to buy, you can download PDF files of individual chapters at www.lenellis.com/books.  An abstract is below. 

Abstract

The assimilation of humans into machines, once science fiction, is a well advanced reality today.  Each of us has virtual versions inside the two great machines of the late modern age.  In the datascape, the vast array of databases in which the details of our daily lives are recorded and analyzed, we appear as profiles.  In cyberspace, the global network of computers in which everyone can connect with everyone, we appear as personas.  Both are part human.  We continually update both machines, passively and actively, and, as we do, our simulacra change in tandem.  Both are part machine.  The profile is a probabilistic portrait, conjured up by others to inform their decision making; it’s an informational output.  The persona is a pattern of connections, created as we present ourselves to and interact with others; it’s a network effect.  Drawing upon humans in near real time but manifested inside machines, neither looks like the continuous, whole and bounded self of the modern tradition.  Rather, these hybrid entities are contingent, relative and open.  Silicon Simulacra describes how these two semblances come to be, how each represents us and what opportunities and challenges each poses and suggests they are the post-human forms of humans assimilated into these machine worlds

Internet Naysayer Doesn’t Go Far Enough

In the “The Dangers of Web Tracking” (August 6 Wall Street Journal) Internet skeptic/critic Nicholas Carr describes how we are and could be tracked, profiled and identified from our online activities.  But the dangers he cites are flimsy and he seems to shy away from naming the danger that concerns him.

Carr offers three dangers. 

The first is crime, e.g.,  identity theft and the frauds enabling by the theft.  Yes, the data generated by our online activities add to the data generted by our use of credit cards, store cards, toll tags, catalogs, warranties, etc., but theft and fraud have no special tie to online tracking.

Second, prediction can blur into manipulation but Carr doesn’t define the latter   If he means behavior-based persuasion, that’s as old as the hills. Every salesman listens to what customers say with their mouths and with their eyes, heads, shoulders, hands and feet and then adjusts the sales pitch accordingly.  Most persuasion encounters are feedback-governed interactions and until recently a task for humans.  In the field of human-computer interaction (HCI), two questions gained a lot of attention early on: could computers teach and could computers persuade?  Marketers are figuring  out the latter in their workaday activities; in specific a lot of interactive design involves mapping out diverse consumers’ different decision paths en route to a purchase and then laying in at each step along  the path the content and/or tools that will lead the customer onward.  This is SOP.  But even ”persuasion” may be too strong a term.  Many designers say instead that they’re helping the customer buy.  If something is wrong with being persuiaded by a machine, I’d like to know what that is.  Labelling it manipulation doesn’t help.   

Third and his greatest danger is the chilling effect of surveillance.  He writes, ” When we feel that we’re always being watched, we begin to lose our sense of self-reliance and free will and, along with it, our individuality.”   It is certainly true that the machine does not care about us as individuals.   Rather we are persons in a population who can be differentiated into groups, some of which are more likely than others to respond to certain persuasions.  Whether the continuous surveillance of our activities for the purpose of parsing us into probabalistic groups erodes our self-reliance, free will and individuality — or even the sense of same — is arguable at best.

Earlier in the piece Carr hints at the danger that I think he actually fears:  specifically, that government could identify those whom it considers opponents.  The headline on the WSJ’s online edition tried to make the point.  This is not paranoia.   To the contrary it’s an axiomatic truth. 

All governments, everywhere and always, have not just a potential but an actual tendency to encroach on the rights of those they are to protect.  Many governments have gone too far; they can and should be prevented and it requires the level of citizen awareness and vigilance that Carr calls for.    Americans don’t like to think that our government could take such a turn, but it did and recently.  Members of the “greatest generation” remember the 1950s Red Scare led by Joseph McCarthy; baby boomers in the anti-war movement can testify to egregious overreaching by both police and FBI.    

If we’re going to deal with the dangers of web tracking, we need to be more forthright in naming them.

WSJ’s Flawed Exposé of User Tracking

There’s an old PR saying—“Never argue with someone who buys ink by the barrel and paper by the ton.”   So, I won’t post this comment on the Wall Street Journal’s web site but will share it here.

Its research on user-tracking software, “The Web’s New Gold Mine: Your Secrets,” reported in three full pages of its July 31-August 1, 2010 weekend edition, is solid on facts but wrong on the one fact upon which its argument depends.  After introducing the types of user behavior collected by cookies, Flash cookies and beacons, the fifth paragraph asserts that this data is packaged into profiles about individuals.   This is wrong as a matter of fact.

Individual-level data are stored in what’s called a record.  Database administrators perform certain hygiene procedures on records and they’re continually updated with fresh individual-level data, but records as such are raw material; they just sit there until someone queries the database.   When that happens, the software scans the data inside each record and sorts the records into groups that comform in greater and lesser degrees to the query.  The resulting group portraits are profiles.  

Typically, the software is designed so that profiles express a probability.  Measuring the variability of individuals on one or more attributes (the raw material), it differentiates these persons as more likely than those persons (the profiles) to behave in the way desired by a business marketer or a government administrator (the query).  This statistical differentiation then becomes the basis for real-world discrimination: treating these persons differently from those persons, in the service of business profits or government efficiency.  Social statistics at birth and all its descendents since including user-tracking software parse populations into probabilistic groups.  The method cannot say—and does not want to say—anything at all about individuals as such. 

That’s why consumers, despite telling pollsters that they are “concerned” or “very concerned” about online privacy, don’t use the privacy protecting tools that have long been available.  They know that this surveillance does not threaten them as individuals.  Scare-mongering about privacy by the media and activists only perpetuates the belief that individuals are important to business and government.  Fortunately, we aren’t.

Word Map App from Thinkmap

The Visual Thesaurus (R) is developed and published by Thinkmap Inc., nee Plumb Design established in 1997.   From day one, a great data visualization app and the novel use is worth a look here —  http://www.visualthesaurus.com/howitworks/.  Re the biz model, we’ll  see.

Housing Meltdown and Data Breaches Have Same Flaw

The housing meltdown and data breaches share the same fundamental flaw: risk and reward are not connected.

In the housing market two Government Sponsored Enterprises (GSE) — Fannie Mae and Freddie Mac (Federal National Mortgage Association and Federal Home Loan Mortgage Corporation) — hold half of all mortgages. Their business — buying mortgages and then issuing mortgage-backed securities — succeeded in many ways and in large part because everyone believed the US government stood behind these loans in case of default. With risk transferred to the government, Fannie and Freddie focused solely on reward, making profits by accepting debt from borrowers whom I wouldn’t trust to hold my groceries, and when these subprime and Alt-A mortgage debtors defaulted, triggering the larger meltdown, the Treasury—that would be you and me, bucko—had to step in.

Data security has the same problem. Publicly disclosed data breaches are rising rapidly. According to the Identity Theft Resource Center, there were 449 disclosed breaches by August 22 compared with a 446 total for all of 2007. When these breaches occur, the harm falls on the consumer — that would be you and me, bucko – but not on the companies responsible. Some 44 states do require them to disclose data breaches but there are loopholes, e.g., only business in certain industries, only if a business suspects the data will be used to commit fraud etc. and there’s rarely a penalty as long as the breach is disclosed. In fact, there’s rarely a penalty for not disclosing – even though that’s a crime – and even when there is a penalty it’s minor — $10,000 in Arizona, for example. The only way to curb data breaches is to connect risk to reward – either allow individuals to hold businesses liable in court or introduce regulations that would impose stronger criminal and civil penalties

The threat of risk disciplines the pursuit of reward. Without risk, someone is certain to get screwed — that would be you and me, bucko.

W.H. Auden on Datamining

The Unknown Citizen
W.H. Auden

 

He was found by the Bureau of Statistics to be
One against whom there was no official complaint,
And all the reports on his conduct agree
That, in the modern sense of an old-fashioned word, he was a saint,
For in everything he did he served the Greater Community.
Except for the War till the day he retired
He worked in a factory and never got fired,
But satisfied his employers, Fudge Motors Inc.
Yet he wasn’t a scab or odd in his views,
For his Union reports that he paid his dues,
(Our report on his Union shows it was sound)
And our Social Psychology workers found
That he was popular with his mates and liked a drink.
The Press are convinced that he bought a paper every day
And that his reactions to advertisements were normal in every way.
Policies taken out in his name prove that he was fully insured,
And his Health-card shows he was once in hospital but left it cured.
Both Producers Research and High-Grade Living declare
He was fully sensible to the advantages of the Instalment Plan
And had everything necessary to the Modern Man,
A phonograph, a radio, a car and a frigidaire.
Our researchers into Public Opinion are content
That he held the proper opinions for the time of year;
When there was peace, he was for peace:  when there was war, he went.
He was married and added five children to the population,
Which our Eugenist says was the right number for a parent of his generation.
And our teachers report that he never interfered with their education.
Was he free? Was he happy? The question is absurd:
Had anything been wrong, we should certainly have heard.

The Devil in the Data?

People get creeped out by data but usually can’t explain why. One reason is that the devil doesn’t lie in the data but in the algorithms–the rules that process the data.

These rules rule. They sit atop the data and crunch it into something meaningful. They could be arguable and could lead to errors of various types but, largely invisible, they are beyond critique.

Fortunately, the Federal Reserve Bank of Boston recently issued a report on redlining by credit-card companies that offers an excellent example of exactly what makes people ill at ease with the machinations of data.

The top-line finding is that credit-card companies discriminate among customers, offering higher credit lines to similarly qualified residents of White neighborhoods than to residents of Black neighborhoods.

“After controlling for the influence of such other place-specific factors as crime, housing vacancy rates, and general population demographics, the paper finds qualitatively large differences in the amount of credit offered to similarly qualified applicants living in Black versus White areas.”

Obviously, having less credit available is a disadvantage but that’s only the half of it.  The other half is a rule that compounds the problem.

In determining an individual’s credit score, the credit-card companies’ rule compares the amount borrowed to the amount available. The smaller this ratio, the higher the credit score. So, two individuals with the same loan and same history of timely repayment will get different credit scores based on the different amounts of credit that were available to them—a factor about which they can do nothing.

There are numerous problems here. Racism is one.  That is a particular version of another and more general problem: the ethics of applying a group-based parameter—here, neighborhood-based credit lines—to an individual’s profile—here, his or her credit-worthiness. But what creeps people out is their “sense” that calculations, judgments and decisions about them are being made behind the scenes in mysterious ways that could be damaging and over which they are powerless.

This concern is are neither paranoid not Luddite. People are right to be concerned and they’re concerned for the right reasons. There’s more to the datascape than data. Over and above data are the invisible rules that determine what data ultimately mean.

 Unfortunately, these conerns are being addressed by politicians in a slew of proposed laws to regulate Internet advertising.  IAB CEO Randall Rothenberg offers a comprehensive and well grounded critique of these ill-conceived initiatives.   But fending off government interference in the marketplace will not address the root problem.   That would require the data-crunching community–public and private–to introduce and examine their practices through the lens of ethics in addition to the lens of optimization.  

 

If 1 million seconds elapse in 11 days, 1 billion seconds…

Millions and billions are the units in which much of modern life is measured.  Unfortunately, humans have a hard time grasping large numbers and an even harder time grasping the relative size of large numbers.  Analogies help and here’s one I hope you’ll remember.

1 million seconds elapse in 11 days.
1 billion seconds elapse in 32 years.

Puffery and Propaganda

Review:  David Weinberger, Everything is Miscellaneous: The Power of the New Digital Disorder

Don’t be bamboozled. This book is puffery on the outside and propaganda within.

A booster of Internet-enabled consumer-generated content, Weinberger hails all its new modes—social networks, tagging, folksonomies, wikis, mash-ups etc.—to argue that the Internet enables new bottom-up ways of organizing and creating knowledge. His two theses are irrefutable. First, information is getting chopped into smaller and smaller pieces. Second, the organization of information has escaped physical world constraints and control by experts and their institutions. The net result: any one can now organize information any way one wants to suit one’s own purposes and do so at any time for one’s immediate use. From this potential Weinberger foresees though rose-colored glasses an ever-changing array of “useful, powerful and beautiful ways to make sense of our world.” Excuse me but such a view has huge problems.

As Weinberger points out, the best grouping and sorting of information into knowledge is done by small groups not only online but also offline, as his own examples demonstrate. The Internet brings nothing special to this process. To be sure, the Internet does make it much easier to form groups and very much easier to form groups that include strangers and groups devoted to narrow topics. But a vast increase in quantity does not necessarily yield any improvement in quality, and although Weinberger is always careful to label such an outcome as “potential,” I’ve yet to see any marked change in our ability to make sense of the world.

Nor is the appetite and aptitude for this activity suited to all. As any teacher who ever graded term papers will confirm, the ability to cut-and-paste information and turn it into knowledge is not innate. Even after some schooling, such a rudimentary skill is very far from universal. It is essential for the new occupational stratum of Information Age employees who manipulate words and numbers for their livings, but such “knowledge workers” are not Everyman. Presenting the abilities and interests of this or any other elite as those of everyone is the essence of ideology and shouldn’t fool anyone.

Nor should anyone accept, much less applaud, Weinberger’s cavalier dismissal of discipline and authority. He argues that knowledge should not have any shape and that deciding what to believe is now our burden. His laissez-faire alternative would have John and Jane Doe fend for themselves, finding and creating meaning from and among each other—link by link, tag by tag. Fortunately, this cannot happen because humans can only make sense of things, especially disconnected facts, through the frameworks of our inherited arts and sciences. Nor should it happen. The casual indifference to several millennia of human efforts to make sense of the world only leaves John and Jane in the post-modern void, unarmed and aimless.

There’s a lot to be said in favor of user-generated content but to suggest that the fragmentation of knowledge and the cacophony of cyberspace will improve our ability to make sense and meaning of life is cybertopian bunkum.