13-10-2018

SCIENCE & TECH
SATUrDAY, OcTOBer 13, 2018
5
Tech’s gender bias nothing new
Marie Hicks
A recent report revealed Amazon's AI recruiting
technology developed a bias against women because it was
trained predominantly on men's resumes. Although
Amazon shut the project down, this kind of mechanized
sexism is common and growing - and the problem isn't
limited to AI mishaps.
Facebook allows the targeting of job ads by gender,
resulting in discrimination in online job advertisements
for traditionally male-dominated jobs from construction
to policing. The practice has long been illegal in traditional
print media - but Facebook's targeting tools encourage it.
Not only can this affect whether women and non-binary
people can see ads; it also affects male job-seekers who are
older and therefore viewed as less desirable by many
employers. Facebook has come under fire for illegal
advertising practices in the past: notably, it scrapped
thousands of microtargeting categories after a 2016
ProPublica report showed how it allowed racial
discrimination in housing ads.
The platform has repeatedly refused to take
responsibility for what people do on it, echoing the
behavior of other Silicon Valley companies. Gendered and
racialized harassment online goes largely unchecked.
Likewise, Google's YouTube has come under fire for
algorithms that appear to push radicalizing far-right
content onto casual viewers, while Google itself has faced
accusations that its image search and autocomplete
features rely on and strengthen racist and sexist
stereotypes.
As online platforms strip away civil rights protections
intended to correct biases in earlier forms of
communication, it serves as an example of the dangerous
tendency of our current, and supposedly progressive,
technologies to re-create discriminatory patterns of the
past. Indeed, these problems fit a pattern in the long
trajectory of the history of technology.
Today, jobs in computing, if advertised on Facebook,
would likely be targeted to men because these jobs are
located in an already male-dominated field. In the early
days of electronic computing, however, the work was
strongly associated with women. It was feminized because
it was seen as deskilled and unimportant. This quickly
began to change as computers became indispensable in all
areas of government and industry. Once it became clear
that those who knew how to use them would have great
power and influence, female programmers lost out despite
having all the requisite skills. Britain's computerization is
a cautionary tale: women were repeatedly and
progressively denied promotions or boxed out of their
jobs, particularly when they married or had children.
Top executives of Facebook, Amazon and Alphabets Inc during a meeting at Trump Tower.
Photo: Shannon Stapleton
When they left, they were replaced by men. This created
disastrous labor shortages that ultimately forced Britain's
decline as a computing superpower.
Women continued to program, but they had to do it
without the support of major institutions. One example
was the entrepreneur Stephanie "Steve" Shirley, who used
a masculine nickname to sidestep sexism. Shirley started a
freelance programming company with an explicitly
feminist business model after finding herself unable to
advance in government and industry. She employed
hundreds of other women who had similarly had to leave
the workforce. Shirley gave these women an opportunity
to use their skills in the service of the nation's economy by
giving them the option to work from home, filling some of
the gaps left by this exodus of trained computer
professionals from full-time computing work.
Shirley's business, built on women's labor and
expertise, went on to become a multimillion-dollar
corporation that did mission-critical programming for
government and private industry. As the government
scrambled for male computing talent, for instance, a team
of her female programmers, led by Ann Moffatt,
programmed the black box for the Concorde jet. As
Shirley's business flourished, many other companies and
even the British government itself suffered for lack of
programming talent.
The irony is that this shortage had been intentionally
engineered by the refusal to continue to employ female
technologists in these newly prestigious jobs. Throughout
history, when jobs are seen as more important, or are
better paid, women are squeezed out - hence the need for
protective legislation that ensures equality of opportunity
in hiring and job advertisements.
In computing today, a field that claims to value diversity,
engineers at Facebook and other companies are building
tools that rollback the advances of women in the
workforce, as the industry undoes the civil rights
protections enacted to ensure that what happened in early
computing does not happen again.
When industries ignore their pasts, they tend not only
to repeat previous mistakes, but also to worsen current
problems. Silicon Valley's gender problems are well
known, and despite companies' claims that they are
trying to address the problem, progress has been slow
and uneven. This is not surprising when we consider the
context. Although the industry is facing a reckoning
today, for decades the stories that we told about
computing technology focused on inexorable success,
rather than taking seriously the possibility that our new
technologies were failing us. High technology became
virtually synonymous with progress and the greater
application of computing to all manner of social problems
was seen as a good in and of itself. As a result, we are
largely blind to the errors of the past. We fail to see the
problems in our present and the reasons behind them
because we are too accustomed to seeing computing as a
success story.
The refusal to talk about computing's failures in the past
has not served us, or present-day computing, well. Rather,
it has hidden problems that have plagued the field since its
inception. Facebook's discriminatory practices towards
female users in everything from job advertisements to
harassment can be traced back to its predecessor, the beta
site set up by Mark Zuckerberg while at Harvard that stole
female undergrads' pictures from internal Harvard
servers. The site, known as Facemash, objectified the
women for an audience invited to rate their relative
attractiveness. When we consider Facebook's current
problems in this light, they not only seem less surprising
but also potentially more solvable.
Lessons like this are critical today because high
technology has an outsize effect on every aspect of our
daily lives, and it is also, in many ways, steadily moving us
back towards a past that we thought we had forgotten.
Much of the anti-racist and anti-sexist legislation of the
20th century has been invisibly rolled back by tech
infrastructures that invite users to see their online actions
as unmoored from real life - whether in the realm of hate
speech or job advertisements.
Strong representation of women in the labor market is
key, historically and today, for women to be able to assert
their rights in all aspects of their lives. Companies like
Facebook cannot be allowed to divide and conquer by
gender, race, sexuality, age, disability, or any other
number of categories people have fought to protect by law
as deserving of equal rights.
Weaponized AI enabling
perpetual wars
A technician checks a server in a data centre.
Do DWeb programs use as much
energy as cloud-based services?
Jack Schofield
The main aim of the decentralised web
(DWeb) is to remove the power of
centralised "gatekeepers" such as
Facebook and Google, who hoover up
the world's data and monetise it by
selling advertising. It reminds me of the
original concept of the web, where every
computer would be both a client and a
server, sharing information on a more or
less equal basis. Of course, that is not
how real life works. What actually
happens is that you get a power law
distribution with a few large entities and
a long tail of small ones.
As Clay Shirky wrote in 2003: "In
systems where many people are free to
choose between many options, a small
subset of the whole will get a
disproportionate amount of traffic (or
attention, or income), even if no
members of the system actively work
towards such an outcome. This has
nothing to do with moral weakness,
selling out, or any other psychological
explanation. The very act of choosing,
spread widely enough and freely
enough, creates a power law
distribution."
The web still has plenty of variety, but
almost everyone is familiar with one
giant search engine, one giant retailer,
one giant auction site, one giant social
network, one giant encyclopaedia, and
so on. Indeed, there is only one giant
internet where there used to be dozens of
competing networks using many
different protocols.
Obviously, it would be better if we all
agreed these things in advance, based on
open standards. However, people vote
with their wallets, and competition
results in de facto standards instead of
de jure ones. Examples include
Microsoft Windows, Google Search and
Facebook. Each triumphed in a
competitive marketplace. I am not
saying this is the ideal solution, just that,
in most cases, it's inevitable.
One of the problems with returning to
a decentralised web is that the internet is
no longer decentralised. It has been
redesigned around giant server farms,
high-speed pipes and content delivery
networks. It looks increasingly like a
broadband television network because
that is what it actually does most of the
time.
Today's web is being optimised for the
delivery of Netflix movies, BBC
programmes on iPlayer, Spotify music,
live streams of every major sporting
event, and so on. You can upload your
own live streams but communications
are asynchronous: your downloads are
much faster, and much more reliable,
than your uploads. It's really easy to
watch 1TB of movies but an exercise in
frustration trying to upload a 1TB harddrive
backup.
If you really want to save energy and
internet resources, stop streaming stuff.
Broadcast TV and radio can reach tens of
millions of people, and adding another
million adds relatively little in the way of
extra power consumption. There is
school of thought that it is better for the
environment to use CDs or DVDs for
albums or films you go back to again and
again, or you could at least use digital
files stored on your PC or smartphone.
Photo: Juice Images
And rather than using Graphite to
replace Google Docs or Microsoft Office,
just use a word processor offline. If you
run Windows, you already have a text
editor (Notepad) and a simple word
processor (WordPad), and there are
plenty of free alternatives. That will
reduce global energy use and increase
your privacy.
It's really simple. If you don't want
Google to read your documents, don't
write your documents on Google's
computers. And if you don't want cloud
servers using energy, don't use the cloud.
Companies such as Amazon AWS,
Microsoft and Google are covering the
world with server farms to make
information more easily available. That's
harder to do with real distributed
systems because the thousands or
millions of separate computers may be
turned off or otherwise unavailable
when you need the data they are storing.
Worse, unless it's replicated, you could
lose data.
It's true that server farms consume
an ever-growing amount of electricity,
much of it used for cooling purposes.
However, the cost is a powerful
incentive for operators to use cheaper
renewables, such as solar panels, and
to reduce their power consumption in
other ways. For example, Facebook
has built a data centre in the north of
Sweden where the air is freezing cold,
while Microsoft is experimenting with
underwater data centres that are easier
to cool. Microsoft is also sponsoring
tree planting in Ireland as part of its
commitment to becoming carbon
neutral.
Ben Tarnoff
Last month marked the 17th
anniversary of 9/11. With it came a new
milestone: we've been in Afghanistan
for so long that someone born after the
attacks is now old enough to go fight
there. They can also serve in the six
other places where we're officially at
war, not to mention the 133 countries
where special operations forces have
conducted missions in just the first half
of 2018.
The wars of 9/11 continue, with no
end in sight. Now, the Pentagon is
investing heavily in technologies that
will intensify them. By embracing the
latest tools that the tech industry has to
offer, the US military is creating a more
automated form of warfare - one that
will greatly increase its capacity to wage
war everywhere forever.
On Friday, the defense department
closes the bidding period for one of the
biggest technology contracts in its
history: the Joint Enterprise Defense
Infrastructure (Jedi). Jedi is an
ambitious project to build a cloud
computing system that serves US forces
all over the world, from analysts behind
a desk in Virginia to soldiers on patrol in
Niger. The contract is worth as much as
$10bn over 10 years, which is why big
tech companies are fighting hard to win
it. (Not Google, however, where a
pressure campaign by workers forced
management to drop out of the
running.)
At first glance, Jedi might look like
just another IT modernization project.
Government IT tends to run a fair
distance behind Silicon Valley, even in a
place as lavishly funded as the
Pentagon. With some 3.4 million users
and 4 million devices, the defense
department's digital footprint is
immense. Moving even a portion of its
workloads to a cloud provider such as
Amazon will no doubt improve
efficiency.
But the real force driving Jedi is the
desire to weaponize AI - what the
defense department has begun calling
"algorithmic warfare". By pooling the
military's data into a modern cloud
platform, and using the machinelearning
services that such platforms
provide to analyze that data, Jedi will
help the Pentagon realize its AI
ambitions.
The scale of those ambitions has
grown increasingly clear in recent
months. In June, the Pentagon
established the Joint Artificial
Intelligence Center (JAIC), which will
oversee the roughly 600 AI projects
currently under way across the
department at a planned cost of $1.7bn.
And in September, the Defense
Advanced Research Projects Agency
(Darpa), the Pentagon's storied R&D
wing, announced it would be investing
up to $2bn over the next five years into
AI weapons research.
So far, the reporting on the Pentagon's
AI spending spree has largely focused
on the prospect of autonomous
weapons - Terminator-style killer
robots that mow people down without
any input from a human operator. This
is indeed a frightening near-future
scenario, and a global ban on
autonomous weaponry of the kind
sought by the Campaign to Stop Killer
Robots is absolutely essential.
But AI has already begun rewiring
warfare, even if it hasn't (yet) taken the
form of literal Terminators. There are
less cinematic but equally scary ways to
weaponize AI. You don't need
algorithms pulling the trigger for
algorithms to play an extremely
dangerous role.
To understand that role, it helps to
understand the particular difficulties
posed by the forever war. The killing
itself isn't particularly difficult. With a
military budget larger than that of
China, Russia, Saudi Arabia, India,
France, Britain and Japan combined,
and some 800 bases around the world,
the US has an abundance of firepower
and an unparalleled ability to deploy
that firepower anywhere on the planet.
The US military knows how to kill.
The harder part is figuring out whom to
kill. In a more traditional war, you
simply kill the enemy. But who is the
enemy in a conflict with no national
boundaries, no fixed battlefields, and no
conventional adversaries?
This is the perennial question of the
forever war. It is also a key feature of its
design. The vagueness of the enemy is
what has enabled the conflict to
continue for nearly two decades and to
expand to more than 70 countries - a
boon to the contractors, bureaucrats
and politicians who make their living
from US militarism. If war is a racket, in
the words of marine legend Smedley
Butler, the forever war is one the longest
cons yet.
Automation has greatly increased US Military's capacity to wage war everywhere forever.
Photo: Getty