Friday, October 31, 2014

Big Brother2.0 debate/conversazione, Lady Mitchell Hall, Nov 1, 11-12.30

I'm going to discuss things from a technical (computing) perspective, but with a modicum of social science

1. The canard: "If you've got nothing to hide, you have nothing to fear" needs debunking
(cue visual duck being thrown out of a hammock:)

This oft-repeated statement misses various important features of the way the world and people work:
Firstly, we all hide things all the time - the reasons are many:

some things are not finished, and need further work before they are presentable - 
sometimes, that is our own selves -our half baked opinions.

some things are hurtful to some people, but not to others

we present ourselves differently to different people - our kin, our close friends, our colleagues, our acquaintances and people that we encounter - all are given different levels of trust, because there are different levels of shared experience (and many other reasons). Context matters

Surveillance is toxic. It reduces everyone's choice of behaviour to that which is acceptable to everyone else. For all time. 

2. Because of the change of context over time (we grow older, our social network varies, the world changes, new stuff gets discovered, people forget stuff), we need to control aspects of information about ourselves as seen by others - indeed, we need to have obsolete data removed from their view

The statement that this is "censorship" is false. It is about a generalization of the "public right to know"

In general, the "public" is a set of people who we can send information to - e.g. my family, my friends, google, facebook, Sky TV, Isil, or GCHQ or the NSA or the Polis. Most of these, most the time, do not have a "right" to know. this is obviously false. I have a right to tell or not. I can judge my context.

e.g. Wes Hardaker, en route from san Francisco to Vancouver in SFO airport tweets to his partner at 5am "this airport is so dead" - the NSA might think he's a terrorist. he isn't - he means that the airport is really quiet (its 5am, after all).

e.g. Euan Blair on his 18th birthday gets drunk and is found /photographed in a gutter. before he was 18 it was no-one's business (he's a minor)- after he's 18, he's an adult -t he fact that his father is prime minister isn't relevant. the fact that many of the journalists covering the story are functional alcoholics and hipocrites is of no more interest, either, even if it is deliciously ironic.

e.g. mark thatcher gets lost in a rally drive across the sahara isn't specially interesting - see above. His mother isn't responsible for her 25yr old son's poor navigation skills doesn't reflect on her free market dogma or handbagging skills

3.with a suitable combination of technology (tracking content using DRM just as music and movie companies do, but on behalf of the citizen) we can tell if people send our data further than we wish, and law (data protection law, esp. in Germany - mainly because over time, the experience of the Stasi surveillance state rammed home why you really should care about this) we can 
catch bad people, fine them, put them in jail and (hopefully) make people think about whether they should inappropriately gossip - we can also age and remove from sight data that is no longer relevant (criminal records for crimes that the perp has rehabilitated, health records of no public interest, financial info that is out of date). THis is no less true of trivia (my birthdate is not necessary for buying a drink, just the fact that I am over age X...)

4. Enforcement ideally should be social, but should include suitable independent organisations - perhaps a new Estate (the first virtual estate)

5. GCHQ (and the NSA etc) are in no special privileged relation to most people in regards the above.
We need to incent them to do their job right. expensive surveillance is not a substitute for good old fashioned Humint....

6. Google (facebook, NHS care data/Price Waterhouse) aren't exempt either

7. we need law with sharper teeth, because of the heavily asymmetric power held by agencies named above compared with the individual

8. data,just because it can be copied without error, is not necessarily true in the first place. and it can become false (law change, for example). recall by humans is revisionist, because context changes. Data without context is inherently false

9. Every decoding is an encoding (Maurice Zapp, Small World, by david lodge).

10. If you don't by this, give me all your keys and all your passwords.

There's a lot of background work to this, but I'm assuming the audience probably won't want bell, book, candle and footnotes:-)

My 10 cents

stephen farrell quite (59 mins in):

terrorism, evidence etc report card:

CATO report on costs of counter terrorism compared to what:

John Naughton's notes are now available too

Friday, October 24, 2014

Not wrong to be forgotten

A few more points after the illuminating debate at the Cambridge Union Society last night (23.10.14).

i) so people are still hung up on the notion that,
because the internet/web/cloud can copy/store things so cheaply,
then deletion is impossible -

first off, that isn't necessarily so,
but even assuming it is,
its tantamount to arguing that
because we can now 3D print guns, we don't have to bother with gun control law...

- in fact, if there is a reason to remove data from publication
(its required by law - because its  defamatory or wrong,
or its required by ethics because the use is no longer relevant,
or its required by emotions - it would hurt someone's feelings again and again)
then the imperative is to embed the right in a law so that we incent people to withold their copies
(which is not the same as _erasure_ but is quite possible, and easily enforced

keeping a copy of my ex-partners photos private is fine (i do) -
re-broadcasting them 35 years later is not.

ii)  preventing search producing a hit on content from i) isn't pointless -
its just a weaker version of stopping people re-publishing stuff
the european court finding was blurry on this
because of the definitions of stakeholders in current law. poor.

iii) the argument that there are many search engines and meta-search,
and that the internet "routes around damage"
so people will setup sites like "blockedbygoogle"
minutes after something is not returned by search, is irrelevant -

if we get the law right, and the right put into action, then technology can work -

scale out of distributing rules for filtering spam (e.g. see how ApamAssinsin shares filter rules, see how ingress filtering and BGP policy checks work, see how legal censorship works etc etc and see how collaborative filtering and recommendation engines work), and
other tricks works vert well (no, not 100% -but we aren't arguing 100% -
we're arguing shifting the balance so
a) people think a bit about this,
b) law that gives victims teeth, so deliberate flouting
earns the perpetrator fines/criminal records, public opprobrium etc

personal clouds, with personal control of flow of data through cloud, and
making the relationships between people and between people and organsations
(whether government or corporate)
more symmetric in terms of control of flow and lifetime of data)
is essential

people get confused about "google do x" or "person re-tweets y" -
the size of the agencies shouldn't
matter - anyone can leverage the internet to scale out cheap copying of data to many,
but should they?, is the question to ask?,
to whom? who is the entitled audience, now? or later? -
see next points:

iv) people are confused that this is censorship -
certainly currently the legal framework, and its "enforcement" by Google is a form of censorship,
because the "right to know" (i.e. public interest)
isn't being tested by an organisation that _represents_ public interest,
in any accountable way - sure, that's a bug, but we can fix that too - that doesn't mean
the right to be forgotten is wrong either

popular journalism (especially in recent years) has not helped :-)

Wednesday, October 15, 2014

Right to be forgotten & trolls & hobbits & elves

A two-part blog today

1. Right to be forgotten.
This is a very bad phrase. The right should be to not be reminded (as much).
The false dichotomy is just that - the binary decision between "remember and not remember".
Neither humans nor technology remember things accurately.

a) Every encoding is a decoding (Maurice Zapp, of Euphoric State University, in David Lodge's fictional Small World) - recall is revisionist for humans. It is also revisionist in technology - you may decide to do an audit (who read the file, when ? these are often recorded or appended to meta data) - where were the people looking at the data? why. Should we sort the results of a search to show more recently accessed data higher? this will affect others. Just as it does with your own grey matter!
b) We need to forget things - its part of coping with trauma, even just at the level of embarassment
c) We evolve (and we all evolve) so stuff we did that they agreed with then, we and they may not agree with anymore. Why should we be faulted now for what was ok back then?

Delete, by Viktor Mayer-Schönberger

Better solutions - social nets should use interaction/activity based weights to decide what is recalled easily - stuff should fade - this copes with public figures naturally (as they interact with world but about their public duties, not their private lives) as well as the private individual online - if I havn't seen you for 25 years, I won't naturally recal lstuff about you - if you show up asking mefor a reference, maybe some stuff will come "flooding back" - this maps also to natural cognitive resource limits that are known about in humans, and implements a nice form of relevance filtering easily. THis also implements "unfriending without tears" on online social media - you just ddidn't interact with someone (like or mention or comment on their timeline) so they slowly fade from memory - many systems already are implementing things a bit like this, but for somewhat misguided reasons - lets fix it - its easy!

2. Trolls
There's a saying from 30 years or more of newsgroups (Usenet, if you go back that far):
"Don't feed the trolls"

So trolling is getting worse - this is not surprising - there's more people in the hills (which have eyes) than ever. These people don't know from Adam (and they are probably eating the apple). Why do they troll - to share in your glory? who knows....but they can because of what I might call Internet Induced Empathic Disorder - they are alienated, and they don't know you are a person - many of them do stuff they wouldn't dream of doing in the same room as you - this is well known. Its worse where there are too many terrible cultural examples (see the Russian Social Media abuse below, but nearer to home, see BBC's :Have your say" comments  or Youtube's hilarious children from hell type comments on videos there (remindes me of the evil devils on rollerblades in the hilarious film, Dogma!)

How to fix? Engage or Ignore? I don't believe the troll's target is the right person - we need to have a social response - peer pressure - show the troll where they sit, surrounded by hobbits and wizards (and elves) - that'll wake them up, or else turn them to stone, come sunrise.

Protest 2.0 - networked negoative consolidation

Wednesday, October 08, 2014

The Lottery of Babel and the Library of Babylon

On every floor of the tower, it is sometimes said, a different language is spoken. once a month, the floor manager buys a ticket for the lottery. if the ticket wins, the language stays the same. otherwise, a new language is chosen, and everyone on that floor must learn it. the chances are rare that the language stays the same for 2 months in a row, let alone a lifetime. however, it is rumoured that sometimes this has happened. Indeed, it is said that there are dizzy heights and possibly sepulchrous depths in the tower where denizens of neighbouring floors have found they spoke the same language as each other for a time. the lottery is administered, of course, by a priesthood, who either understand all the languages, or else have their own universal tongue and each learn the language of a handful of floors. nobody knows which.

few people on each floor are capable of learning a new language every month, and so most people cannot truly understand one another. sometimes, a group get together and try to keep the old tongue alive. these cathars claim too that there are no priests and there is no lottery. they are quickly suppressed.

In the library of babylon, the most sacred works of all religions are kept. the more common books, such as the bible, the torah, the koran, the vedas and so on occur in many versions. the rarer religions, or ones for whom there are few remaining living followers, often have a single copy, or even only a partial segment of the Ur text. the organisation of the library is chaotic. the librarians wanted to organise the text in order of date, but could not agree on whose calendar to employ. then a small group  proposed alphabetizing the entire collection, but were resisted by those who pointed out that there were more ways of writing than there were religions.

the theological experts suggested that there was a tree of religions, not to imply "older" or more "fundamental", but that one could see many similarities in the gods - the indo-european polytheists at least, and the abrahamic religions.

the chief librarian suggested that this was merely the result of errors that crept in during the copying of texts due to the biases of the monks at the time.

as a result, no-one can vouch for the correctness of any of the religious tracts. reform and born-again are thrown together in confusion. there is a ground swell of opinion that the library must be burned, as it represents the ultimate sacrilege. the librarians have all deserted the buildings, which stretch far across the sands to the sea, where the boats full of foreigners are arriving. it will not be long now.

Blog Archive

About Me

My photo
misery me, there is a floccipaucinihilipilification (*) of chronsynclastic infundibuli in these parts and I must therefore refer you to frank zappa instead, and go home