The Magazine of IEEE-Eta Kappa Nu July 2017 - 10

FEATURE
Wise engineering educators have long
emphasized the consequences of errors and
unforeseen characteristics in engineering
designs. For me, it started on my first day as
an undergraduate when I was required to
watch a movie of the 1940 Takoma Narrows
bridge failure. Today, an Internet search on
"engineering failure" will provide a long and
humbling list of examples. With these failures
in mind, it is clear why the codes of ethics of
many engineering professional societies call
on engineers to "hold paramount the health,
safety, and welfare of the public."
The responsibility and challenge of adhering to these
codes must now embrace the rapidly expanding
influence of Artificial Intelligence and Autonomous
Systems (AI/AS) within society today. When machines
can exhibit what may appear as morally based decisions
to its users, the application of ethical methodologies to
the design of technology before it's released becomes
of paramount importance. As Gordon day states, "How
do we gain the benefits of artificial intelligence while
maintaining and respecting the nuances of moral
and ethical principles? Foundational questions such
as these stretch our codes of ethics, and should lead
us to examine the relationships between people and
machines more deeply."

Ethically Aligned Design
In April of 2016, The IEEE Global Initiative for Ethical
Considerations in Artificial Intelligence and Autonomous
Systems was launched with a mandate to provide
a document that could provide technologists with
a pragmatic guide to deal with the pressing ethical
considerations of AI/AS. While it is not a Professional
Code of Ethics, The IEEE Global Initiative has created
the first version of a document called, EthicallyAligned
Design featuring over eighty specific Issues and
Recommendations written by more than one hundred
global thought leaders in AI/AS, ethics, policy, academia
and business. The purpose of The IEEE Global Initiative
is to ensure every technologist is educated, trained, and
empowered to prioritize ethical considerations in the
design and development of autonomous and intelligent
systems. Ethically Aligned Design was created as a
complement to traditional Codes of Ethics to aid and
empower engineers not familiar with these technologies
to increase innovation while diminishing negative
consequences in their work.
10

As many technologies have been transformative in the
past, some may feel new codes or considerations may not
be necessary for AI/AS. Many others, however, feel the
nature of how these technologies affect human agency,
autonomy and intelligence by design justify a deeper
rigor regarding ethical considerations. Konstantions
Karachalios, Managing Director of the IEEE Standards
Association and Member of the IEEE Management
Council noted the following in our interview:
I would compare other technologies from
the past with the DNA of modern technology
landscapes,
whereas
computing
and
algorithms (AI/AS) are becoming their RNA.
AI/AS technologies will be the messengers of
information, transferring it to the right recipient,
interpreting and suggesting courses of action
regarding human choice. It is difficult to imagine
any field of human activity that may escape
such 'benevolent artificial assistants', while still,
as with RNA, we may not be in a position to
fully grasp how AI/AS exactly functions at all
times.
More generally, I see here a risk of loss of
human agency in favor of non-human entities
who 'know better than we know what is best
for us'. One can easily imagine algorithmic
systems deciding about credit worthiness or
admission to universities with minimal human
intervention (if at all). There are also other
major challenges, like explicit or implicit bias in
algorithmic design, transparency problems with
self-learning systems, politically troublesome
asymmetries in agency, use and control of the
data oceans generated, and so on.
When asked how he feels these challenges should be
addressed, Karachalios went on to say the following:
Nothing less than a new wave of enlightenment,
targeted at the implicit beliefs around this
technology, will be necessary to mitigate the
risks and materialize the benefits for humanity.
For IEEE it is not a matter of ambition, it is a
sense of duty that is driving us.
To accomplish this duty, IEEE must grow
beyond its traditional engineering tradition
and make a serious effort to address the not
functional challenges of AI technology, simply
because there is no other actor that could
www.hkn.org //

THE BRIDGE
http://www.hkn.org
Table of Contents for the Digital Edition of The Magazine of IEEE-Eta Kappa Nu July 2017
