apologize or else

ranting - talk radio for the dea[fd]

Tuesday, April 15, 2014

Bibliometrics and Science - Failure to Understand the Basics of the Discovery Process

SO I've read yet another well meaning article on bibliometrics (like the H-Index) and why they might be ok for evaluating groups or sub-disciplines (disagree) but they are definitely not ok to evaluate individuals (agree, but for different reason).

This time, this paper on Bibliometric Indicators of Young Authors in Astrophysics: Can Later Stars be Predicted? hit the twittersphere, hence caught my attention.

Look, all these papers treat the research publications world like some high school statistics project. Ok, why not raise the game a bit.

Let's suppose that scientific discovery is a complex natural phenomenon. Let's suppose there is such a thing as progress :-)

OK so what would the time series of discovery look like? My simple minded hypothesis is that it is (like many other natural processes in a complex world) a self-similar arrival process.
So how do we characterise such a time series? well, it isn't captured in a single statistic like "mean", or even two (mean + variance) - the point of such, essentially fractal structures in time, is that they are characterised by very complex descriptors, and, crucially, prediction is hard - exactly why the weather, and associated phenonomena like flooding, and volcanic eruptions, are hard to predict on an individual basis, although, collectively, we can model broad trends. Surprise surprise (literally and figuratively:)

So science doesn't depend on a random walk in a well structured but sparse or even poisson point random space, where walking faster gets you more results. Nor does success depend on hard work (more sweat, more kudos). While a slightly more random walk might get you an inherently more surprising result, it isn't necessarily going to yield more results. And more work only pays off after the discovery, when you want to present it properly (I am sure history is littered with holes made out of discoveries that were cool, but so badly reported they were ignored and lost).

So predicting the next big discovery by a specific scientist is a bit like saying that a raindrop is going to fall on a particular rain gauge at a particular minute of a specific hour on a special day. OK if you are the bookie setting the odds, but I wouldn't bet on it.

Monday, April 07, 2014

writing like its in the bag

there's this type of informal writing that just missing the goal, the target, like it wants to enagge with you visciously, but the speaker hasn't quite got the peg to hag the metaphor upon, and some words are just de trop, perhaps naive manque.

normally, as if there is such an as if, the problem comes around to framing, err, the problem - sometimes, the audience is abused, left holding their misapprehensions in their handbags, but usually, its just that the writer can't. that is. write. for sh. it stand's to reason, isn;t it, that some people just don't have the gift of the grab-all, the momentricious blue steel glare made of pure glyphtitude, that indietinguished suits you SIR epidemic enthusiasm for exapansionist gestures. oh dear, there i gove again.

Wednesday, March 19, 2014

Towards an Inviolable Internet

Resilience and Privacy - Jon Crowcroft 18.3.2014

Building an inviolable Internet

1. We want generational computing:-

Sublime reason
i) According to Vikram Chandra (geek sublime) there are 300,000 Sanskrit manuscripts
on palm leaves...  vanishing at the rate of several a day due to corrosion

Trivial pursuits...
ii) my life bits  - formats, kids - grandma's degree from st Petersburg 1971 etc


=> Resilience - tolerant to faults/flaws -> Integrity

2. We need information resource availability and correctness "5 nines" & better:

Seriously
i) the internet is my memory bank
my house deeds
my bank
my shopping
utility controls

Trivially
ii) and a source of end of endless pub disputes about facts

=> Availability

3. We have been sold centralised systems (the Cloud)
which aren't incented to do 1 or 2....

i). we can suddenly lose data if a central company goes broke (loss of generation)
ii) our DSL line is down or cell phone coverage is poor...(loss of availability too)

4. We'd like Confidentiality - privacy....(why - see later)

Trivially
i) Central cloud services are tempted to mission creep
targeted adverts and analytics (market research undermine confidentiality
accidental breaches leak massive data sets to the entire world

ii) Government mission creep leads to massive surveillance


Q. Are there alternatives?
A. Yes

1. Encrypt all human related data when stored as well as when transmitted
=> trustworthy encryption software
=> key management complexity
=> can we extend encryption to "safe" processing?

2a. Own the storage - whether in home, pocket or cloud
=> payment if centralised [because central cloud server no longer has analytics/advert revenue)
=> what are costs?
actually, a lot less than you think -
take google+facebook revenue/number of users (conservative (high) estimate
=> 3 euro per month  << internet access bill for broadband or cellular
=> could bundle with network access
=> Problem - central site is still open to
Coercion, Corruption, Connivance with Big Bad Agencies...
(coerce to weaken crypto or reveal keys etc etc), so how to tackle that...next, decentralize:
2b. alternative:
Decentralise - i.e. peer to peer
=> what are the incentives?
actually mutual benefit, but
could pay (much less than 2a, due to lack of need for big servers
soem electricity cost increase to home server (<< 2a)
=> what are risks?
store some other person's bad or embarrassing data
encrypt
=> efficiency
interesting approach is to "code" data so k/n is sufficient to recover all
much lower overhead than central (full) copies
=> Mutually Assured Destruction
Eternity service was envisaged e.g. to store
e.g. BBC + VOA + Al Jazeera on same servers
can't remove one without removing all:)

3. regulatory, legal and economic  control of breaches (sever penalties)
you would have different terms & conditions with central encrypted storage provider
or peers...


4. Ethics:- "You have nothing to hide, so you have nothing to fear"
This statement is nonsense -
The problem is that something on the Internet is not just not hidden,
it is effectively broadcast. This is problematic

a). Social
It is human nature to present different persona to different people
Removing this right is psychologically toxic

People have different points of view - this is normal

b) Personal
We live and learn - we have the right to make mistakes
(even to commit crimes and misdemeanours)
and have them forgotten (mainly) - many such
"embarrassments" -
minor drugs offences, terminated pregnancies
treatment for STDs, depression

c) Government (and law) change

Do you want to go back to the chill of not discussing socialism in McCarthy era USA (the real cold war)?

or civil rights in southern states?

or workers rights in 1930s England?

or if you are right wing libertarian, the removal of coal miners excessive union power under Thatcher?

or the government of South Africa having discussions with  Mandela in prison about handover of power , or of England with IRA about northern Ireland peace agreement?

d) Who polices police?
LovInT incidents

Worse - again, one click leak of all data  -
could release location of abused partners
to dangerous men, or of people under witness protection programmes
or of information that was gathered for intelligence,
but not intended as meeting laws of evidence (i.e. insufficient
for court, but bad enough for newspaper).

Conclusion.

Recent revelations mean that the governments (esp. of UK and US) have
"weaponised the Internet" against civil society.
They have broken the social contract about
what is reasonable to do, and the Internet must be fixed

To do so will mean that it becomes much harder for government agencies to track genuine bad guys - this is their fault - had they stayed within bounds of lawful intercept and civil society's understanding of that, there would have been no need to make the Internet and the Cloud inviolable.

The NSA and GCHQ  have forced that requirement on civil society and will have to work with the consequences. That is, after all, their job.

Saturday, February 22, 2014

I beseech you from the Vowels of Christ, think it possible you may be mistaken about Big Data

A. Just because you already have our data in the bAg, doesn't mean that you shouldn't delete it forthwith and cease and desist messing with our privacy

E. I bEg you, stop beggaring us, by selling our Big Data cheap (we care what you do with care.data - use in the NHS, fine, use for Big Pharma, without and decent cheques and balances, no way)

I. My data isn't bIg - i am small and insignificant - so why include me? havn't you heard of the central limit theorem (and the statute of limitations:) ?

O. All this data, without a purpose is just a bOg, a swamp, a mess - where are your schema?

U.  much of the data is bUg ridden. just because you collected it off of the Internet doesn't tell you much about povenance, validity, accuracy, timeliness, etc

So as with Cromwell (an with apologies to Rimbaud), I beseech you from all the vowels
Bag, Beg, Big, Bof Bug off our data.

Friday, January 17, 2014

mashups and simulations - time travel and robot laws

two ideas from the Free Plotware Foundation today - two for the price of 1

1. in the future, there will be artists who compose mashup installations from pieces of your life using time machines to segment and re-loop, add paradoxes, and alternate paths as they (and maybe you) wish - this will be the most popular art form of the 20th century until it is destroyed accidentally by an A bomb

2. it turns out that we live in a simulation - we knew that. but who knew that god's programmers were playwrights?

basically, the gods give us a bunch of ethics (think "the 10 commandments")
and then ask the playwrights to see what goes wrong in one of two ways:
i) communication failure (comedy)
ii) semantic error in node's code (tragedy)

the reason is that the universe is too short lived to run any verification of the systems of ethics, so they try them out in lots of random settings instead (la cage au folles, the duchesse of malfi, - did you know that in some ethical frameworks, Macbeth is a hilarious comedy?).

Think "Asimov's 3 (nay, 4!) Laws of Robish" only writ on a bigger canvas:)

Wednesday, October 30, 2013

T&C

By the time you have read this,
you will have agreed to the terms and conditions
that I set for dialogue with me,
within me, and without me.

The providers of this,
and other channels of communication
pretend to believe that they have some notional
 control and ownership over things that happen
on the channels they provide.


They are wrong. Historically, ideas trump contracts,
because the social contract they sit within
 is obviously a super-class for any other
 kind of contract that could ever have any currency.

The world in which we live (in) has delivered
 zero-cost copying, so
 information doesn't "want" to be free  - it is free -
people that block its flow are committing crimes against nature.

this doesn't mean we have no privacy -

privacy is the tool for de-tox in a world
when we can't cope with getting our head around differences.

So if you can't deal with the difference
between you & me, between them&us,
then you've invaded my privacy

But if you want to stop me saying stuff,
then you've stolen my voice.

These are my terms and conditions
and if you violate them, you may find.

zero-click enabled. too late, you've agreed.

Wednesday, July 03, 2013

glastonbury 2013

weather verdict: best ever:) only 1 day of rain - 4 of sun! redoculous-  ended wearing sandles shorts/t shirt (don't even blink:)

was seeing steve winwood doin a version of this song,?
somewhat like this arrangment...
http://www.youtube.com/watch?v=eoSn2Y-b6wI
(see this version 28 mins in, from 1969 - scary!)
http://www.youtube.com/watch?v=YfAHsiTHWfQ

Fatou Diawara was even better, as were Chic and rita traore....and the congos
(brief catch of their roots reggae)

james blake and cat power and Bobby Womack -
that made a great sunday nite actually, as it was so BASS centered!?
you couldn't help but _feel_ the low tones - have to say this is something you cannot reproduce on iplayer (you probably don't have a 16*18" speaker system anyhow:) so you had to be there....

beady eye (who were a surprise show) were really quite good....

as were, ben howard and elvis costello and rufus wainwright
and sinead oconor was scorching - shame didn't see her whole act...

didnt think much to vampire weekend (as usual)
but Alt-J were good...
and laura mvula was pretty good (little bit I caught)

liked the Staves a lot
and quite like grrl band Haim, (from LA)
and I liked the last 2 songs from The Heavy!

arctic monkeys were v. good (as expected)?
as was billy bragg

dizzee rascal and publc enemy were ok (if you like that sort of thing)

tom tom club were fun (as expected)

goat were weird and the xx were a bit dull (saw a tiny snippet of them)

shame to miss nick cave, as I think his live act is stellar....

but the stones really really stle the show - especially frm poor Primal Scream
who were lookin like a pale shadow:) - version of midnite rambler and can't you hear me knockin with mick taylor - scary - when sympathy for the devil drumconga intra starts, primals vanish in the dust in the rearview mirror.....

an some people claimed you don't get much for 200 quid - i'd say that its a
bargain!