IEEE Technology and Society Magazine - June 2023 - 34

Commentary
audiences, analyzing customer satisfaction, and
optimizing supply chain operations. This focus in
the areas of AI adoption has, we argue, colored our
very understanding of what " smart " systems are:
economics-driven and individualizing. That is not
to say the use of AI technology in commercial applications
is unwelcome (though in some cases it has
been). Rather, there is a missed opportunity for this
technology in other spaces, and such an imbalance
may bias our assumptions about what is possible or
where the opportunity lies.
So, how could AI show up in digital public spaces?
What could it be used to achieve? And how might
we build that?
In many ways, it has seemed as if we are on the
back foot when it comes to defending rights online.
Indeed, even the common narrative of " defending "
rights, rather than securing or furthering them,
betrays the way many of these platforms have
operated.
There is, of course, a strong and healthy technical
community motivated to present alternative
platforms and respond to emerging threats. For
example, a simple individual choice one can make
is to use Signal instead of WhatsApp, or Mattermost
instead of Slack. As researchers, we can work on projects
that expose unfair and socially harmful biases
in machine-learning models, and we can challenge
design assumptions. Yet, all these acts are essentially
defensive.
Possibilities of artificial intelligence for
digital public spaces
What might positive, creative, socially focused
uses for emerging AI technology look like? If we
could divert some resources to being proactive
instead of defensive, perhaps we would have the
space to create new, interesting, and empowering
things in the public digital realm. Perhaps, we could
design technology-enabled spaces for the purpose of
strengthening the community and empowering community
action instead.
Perhaps, the most important are mechanisms for
self-organization, facilitating deliberative democracy,
and enabling collective action based on common
knowledge. These are examples of the kinds of
things that we are largely missing in the AI that we
conceive of today, certainly in the corporate world.
To have the right to do these requires sovereignty
over space, to secure the space for such purposes,
34
without the need to rely on an alignment with the
owner's commercial interests.
As one example, we are exploring how to design
" smart " systems based on the principles and mechanisms
to create deeper cooperation in neighborhood
communities. The core idea is around building
and sustaining social capital through community-oriented
recommender systems. This is something
some of us have been codesigning with community
housing groups, community leaders, and nurses in
the Black community in Toronto, with the aim of
empowering communities to solve food security
challenges.
There is precedent for this form of technology-enabled
social action: during the COVID pandemic, the
virtual " Auntie Betty " app, developed in the public
interest by The Pamoja Institute created a network
of essential community support resources able to
connect with vulnerable people in meeting health,
wellbeing, and poverty needs, at one the most difficult
of times [15]. This in itself was hugely impactful,
but what quickly became apparent was the structure
and value of the community network that emerged.
Qualitatively speaking, the app's social network
contained a ring of " aunties, " each supporting one
or more subgroups while also mutually supporting
each other. This was not predetermined; it had
emerged as being effective.
Yet, at scale, network issues exist, and people
can " fall through the cracks, " leading to pathologies
such as isolation of individuals, bottlenecks where
individuals are over-relied upon, and weak points
in the network, where the absence of one or two
people could risk the whole community's resilience.
In our ongoing research, we are exploring how to
harness recommender systems to strengthen these
community networks. The core idea is that we do
not simply make recommendations based on what
an individual is perceived to need (or might buy),
but based on what the community needs to thrive,
and, with all the appropriate level of transparency,
explain why such a recommendation is important.
This enables us to build social capital, both to bond
existing networks in more resilient patterns and to
bridge between subcommunities, to provide additional
resources and flexibility when necessary.
The food security challenge of the 21st century
will not be solved by more food banks. Instead, its
solution lies in communities that are empowered
to be able to address and ensure their own food
IEEE Technology and Society Magazine
IEEE Technology and Society Magazine - June 2023

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - June 2023
