Hi, everyone. It is great to be here. Thank you for coming to this talk. If you're here
for the magic show, I'm afraid you have 30 minutes to wait. I'm here to guide us in an
exploration of what we as a community, as open source practitioners, can learn from some
of the most finely tuned and highly performant teams in the world. First responders. Through
the interdisciplinary lens of social network science. So perhaps there is some magic in
this talk. The magic of people working together. My name is Hannah Aubrey. I lead fast forward
at Fastly. Let's save the Internet. In a past life, I was lucky to serve as a study coordinator
of Sonic. No, not the one with the roller skates and the hamburgers. Dang, I knew that
joke wouldn't play in the EU. The science of networks and communities research group. Sonic
advances social network theories, methods and tools to better understand and meet the needs of
diverse communities. They develop cutting edge techniques to study and improve social and
knowledge networks and distributed working groups, online communities, virtual teams, and other
large communities like the one we're all in. I am thrilled and a little bit washing to share that
with the director of Sonic. Professor Nasheer Contractor is here in the audience today. Thank
you for coming, Nash. And my dear friends, if you have any tough questions, please direct them at him.
Let's start with a history reminder. Our earliest ancestors not only had to contend with the same
natural disasters we experienced today, they also had to adapt and survive to nature itself. First,
we became bipedal, freeing our hands to reach and to grasp and also to communicate simply with
each other. Next, we developed complex brains with prefrontal cortexes, our personality centers,
which enabled us to make snap second decisions based not only on external stimuli, but also our
past experiences. Then we developed symbolic language to communicate complex ideas and then
finally tools to take control of and shape our surroundings. So you see what makes us uniquely
human. Actually, what has brought us here together today, the abilities to ponder, convene, reflect,
build, collaborate, and coordinate are not only what make us so special, but also so successful.
So then our tools got a lot better. The first fire pump was invented in Alexandria in the
third century BCE. Unfortunately, it could not save the library, but I digress. As societies and
civilizations began to form, the blast radius of disasters grew. We settled into towns that could
burn down and buildings that earthquakes could topple. And so those smart brains of ours formed
teams whose sole purpose was to patrol and respond to natural and made manmade disasters in the form
of firefighters and police forces. Then societies became more complex. And with that came more
complex disasters, not only fire and flood, but we created monetary systems, banks that could
collapse and food systems that were prone to mass famine, not always for lack of food, but
sometimes for lack of transportation or poor planning. Our close proximity to each other in
cities and long distance cultural exchange made possible by ships brought diseases, colonization,
and war, which ravaged human populations. We think of these ages as dark or undeveloped, but
their responses to such crises were surprisingly neither. In fact, we begin to see thoughtful and
multifaceted disaster response, not only search and rescue or medical aid, but tax relief,
temporary infrastructure, even what we now call refugee camps, providing long term food and shelter
for displaced peoples. In 1493, the Knights Hospitalers shipped doctors and surgeons to the
Greek island of Kos after an earthquake. And so we see some of the first evidence of multiple
different groups or organizations coordinating across disciplines and borders to respond to a
disaster. In the intervening years, we've continued to hone our disaster response strategies.
Humanities impact on this planet has required us to do so. And besides, those prefrontal
cortexes of ours have a lot more data to lean on than our friends the ancient Alexandrians had.
If they knew then what we knew now, maybe they could have saved that library.
You should pull it together. Anyway, today we have entire organizations, governmental bodies,
NGOs, and community groups dedicated to such activities. We have laws by country and internationally
to enshrine basic human rights and ideal responses in crises. And now we're building a new frontier,
a new form of transit. We're creating massive new civilizations, hosted on smallish, inscrutable,
blinky boxes. In this new world, we can't even really see the threats, the crises. We're throwing
people together in a way that's affecting global social structures and people's everyday lives.
Like every form of infrastructure, like most every place where humans gather to live, to work,
to learn, to play. The internet has grown up in an unplanned way. And we're still scrambling to
understand it, to learn from our mistakes, to apply those lessons, to build the best internet,
to build systems that protect people and systems that react when people are harmed.
But don't worry too much. We'll survive these dark ages. Our species has survived every
disaster it's encountered, at least so far. A common organizational structure found in groups
undertaking large-scale operations to solve big, big problems is called a multi-team system.
A system comprised of multiple teams working towards a shared goal. These structures can
be found throughout all sorts of industries, working on all sorts of problems, disaster
response, space exploration, governing humans, building stuff. If you're part of a business
with multiple departments, you're in one. If you attend or work in a university, you're in one.
And if you maintain or contribute to a support or, excuse me, contribute to or support or care
about an open-source project, you're also in such a system. Because no matter what corner of the
internet you occupy or to which technology you contribute, you're working in service of our
shared mission to keep the internet open and free. So what makes up a multi-team system? Within the
subordinate, the superordinate team, the entire system, we have local teams working on local or
proximal goals, which may even be split further into component teams. And directing the subordinate
teams is the leader or perhaps the team of leaders, which shares a global or system goal.
And when you examine these teams using social network analysis, you find common patterns
between successful MTSs. There are many more patterns we could discuss, but let's focus on three.
A plan for coordination paired with frequent, clear communication, highly-performant and resilient
local teams, and finally, empowered and effective leaders who are willing to sacrifice their local
goal in service of the global goal. So before we explore each of these patterns, I want to share
this diagram with you to underscore the importance of these patterns in disaster response. Because
that term disaster response makes such activity sound reactive, doesn't it? But in reality,
the most effective disaster responses begin long before the disaster happens or second best
right after a disaster occurs. So I ask you to bear that in mind through the rest of this talk.
After all, the best time to plant a tree was 10 years ago, and the second best time is today.
First, let's talk about planning, coordination and communication. I don't think I need to talk
about docs too much here. I think the OSS communities know this one quite well. And engineers know
all about retrospectives. Like I mentioned, disaster response begins well before the disaster
occurs. So in terms of coordination and communication, knowing where to turn for help or resources before
a disaster occurs spares valuable time, energy and mental load during a crisis. Effective
communication prevents errors in the field, helps the even distribution of resources and helps us
learn from the mistakes we've made last time so we don't make them again next time. During disasters,
response teams crucially over communicate. They share reports on the situation as it evolves.
They communicate with stakeholders on the ground. And they report changes or progress to make the
best decisions. Leadership and subordinate teams must have the most accurate and up to date information.
Because knowledge sharing fosters a coordinated and collaborative environment.
It reinforces the multi-team system as a single unit, not a set of separate teams.
And because knowledge sharing makes it easier to be flexible and adaptable in rapidly changing
environments. Interestingly, research has found that inter communication, communication between
local teams is more important than intra communication. Communication within the local team to the
success of the whole system. So in fact, there's actually a Goldilocks zone of inter to intra
communication. Local teams should communicate half as much between teams as they communicate
within their own team. Any more inter communication than that and performance declines any less than
that and it declines too. When we talk about the viability of a team, we mean the success of the
team. In moments of disaster or crises, the stakes are life and death. And at the end of the day,
disaster response teams and open source maintainers too, they're people. They have feelings. So viable
teams or successful teams support each other. They lend a hand. They take emotions into account
when making decisions. Viable teams engage in what is called disruption, buffering behaviors,
which is to say change management. They try to anticipate changes that may occur, plan ahead
and invent that some change or disruption occurs. And again, they support each other through those
changes. Viable teams also try to balance performance and resilience because when you work with people
and you're so hell bent on performance, then the team's physical or mental health is at stake.
And the team becomes brittle and the team does not perform well.
So when we see teams that are so, because people do not want to be a part of such a team, right?
So I'll say that again. There's a difference between successful teams and teams that people want
to be a part of. And in the long term, teams that strike the right balance are the ones that are
the most successful. Finally, the most important, the most performant teams strike the right balance
between reinforcing the team's boundaries, which is to say reinforcing the identity or team spirit
of the local team and boundary diminishing behaviors, which reinforce the local team as part of
something larger than the team as part of the whole system. So a little bit of silo is actually good
but not to the extent that teams develop an us versus them mentality,
which brings us to our last assertion today. Empowered and effective leaders.
Strong leaders serve as an ambassador to the team and for the team. Internally, they help teams
understand why the team has a certain goal or is performing some task. Within the system, they
advocate for the team's priorities and points of view. Those are called boundary spanning behaviors.
They make sure that the team has the information it needs, not only the what, but the why of a task
or priority that they understand their own team's priorities. In a disaster response scenario,
times of the essence. Rapid decision making allows teams to quickly assess the situation,
evaluate available options and act promptly to address emerging challenges.
Delays in decision making can lead to missed opportunities, increased risks, and further
escalation of the situation. And as much as we're proud to be a part of our own team, we must recognize
and understand other teams' respect and contribute to their priorities and not be too selfish in
our own focus. That's why a crucial feature of successful multi-team systems, of disaster
response effectiveness, is that local leaders and teams are willing to sacrifice their local goal,
if it means more for the common good. So now that we've immersed ourselves in the theory of
effective multi-team system performance, let's illustrate it with a real-world example.
I recently discovered this amazing YouTube channel. It's called Brick of Mortar. It's all about
infrastructure disasters, ship sinkings, critical failures. It's fascinating. If you're into this
kind of stuff, check it out. You'll never look at bridges or tall buildings the same again.
The sinking of the MV Ferry Wall on April 16, 2014, off the southwestern coast of South Korea,
was a disaster, not only in and of itself, but also of multi-team system performance.
Over 300 people paid the price for these failures with their lives.
On what seemed to be a trip like any other, the ferry suddenly made a series of sharp turns. But
as we know, a disaster such as this starts long before the immediate catalyst. Over the years,
this ferry had been repurposed many times and additions had been made that affected its balance
point. For this trip in particular, the ship had taken on excessive cargo, which compromised the
vessel's stability and made it more susceptible to capsizing. What's more, the ship's crew had
drained the ballast that's water that's kept in a ship to make sure it doesn't sink, to make sure
it's properly balanced. They didn't want it to sit too low in the water, they wanted to be able
to pass inspection knowing they'd taken on way more weight than they were supposed to.
So the communication breakdowns. First, when the ship began to list, the captain refused a
sendage of stress call during the crucial first moments, delaying rescue efforts as the ship began
to sink. He told passengers to go to lower levels of the ship after refusing to tell them anything
about the impending disaster. For crucial moments when they should have been getting on to the deck,
getting ready to be rescued. When he finally sent the distress call and rescue ships came,
they quickly learned that the actual communication infrastructure, the radios that the ship needed
to call the disaster teams were either malfunctioning or were broken. Something had gone wrong with
them. So despite the rescue teams trying to raise the ship's crew on the radio, vital communications
failed during those crucial first moments. So you can see the ferry seawall had no plan for
intra communication in the event of a disaster. They coordinated poorly, not only within their
local team but also with the rescuers. So they failed to inter-communicate with local teams. So
the system, the global team failed. So for the sake of this section, let's quickly divide
the various local teams. The crew is a team, the rescuers are a team, the passengers are a team,
and the South Korean government is a team. What were each of those teams' goals? Passengers
wanted a safe trip. The crew should have wanted to get them there safely but they just wanted to
maximize profit. The rescuers wanted to make it to the site quickly and save as many passengers as
possible. You would think the South Korean government would want to save their people
and prevent such a disaster from happening again, but unfortunately that was not the case.
Their true goal was to save face on the international stage.
We'll talk more about that in a second. Now each of these teams had goals that were in opposition
to another team's goals. And as the circumstances evolved, there was no ability of any of these
teams to shift their priorities or manage this change to negotiate the priorities and evolve.
And each team in the system saw the other team as a detriment to achieving their own goals rather
than as a part of a system, as allies, as individuals worthy of consideration. In fact,
the crew had never received proper safety training. So even if their goals had been aligned, they were
not properly equipped to perform. Now the next example from this horrible tragedy is an example
of leadership, failure, and boundary reinforcing. When rescuers arrived on site, the assembled
parties included Japanese Coast Guard and the US Navy. When a ship sinks, often there will exist
air pockets within the ship. If passengers can find them, they can survive for seven days as long
as they have food. The US Navy or water, pardon me, the US Navy and Japanese Coast Guard and
private citizens too all had the equipment necessary and were on site, the equipment necessary to
conduct such a rescue. But due to South Korea's rigid hierarchical culture and their government's
desire to save face, the teams that had the equipment necessary were not allowed to perform the rescue.
It's an example of unwillingness to sacrifice the local goal and a lack of emotional and really
life support to the passengers who just wanted to survive. In fact, throughout the crucial hours
then days when those high school children that were trapped in that ship could have been saved,
the South Korean government lied to the parents who had assembled to wait for news about their
kids. They said that all the kids had been saved despite that being quite far from the truth.
So what do I hope the open source community will take from this line of scientific inquiry,
from the lessons of the MVC wall? Because folks, this ship is sinking. Our planet's ecosystem is
failing. The climate is changing. I hope when projects, especially leaders, see someone building
something similar to what they're doing, they start to think that other project is an ally.
That other project is an ally, not a competitor. They think, how can we help each other? Not,
how can I win? Or worse yet, how can I sabotage them? I hope maintainers who make the commitment
to serve their community understand the commitment they're making and live up to that responsibility.
Because remember, it's not a commitment you have to make. You can make something, you can choose not
to maintain it, you can choose not to accept issues, change anything about it. But if you make that
choice, I hope you live up to it. And I hope you respect your community and listen to what they need.
I hope BDFLs, benevolent dictators for life, focus more on the benevolent part and less on
the dictator part. I hope we can take better care of each other. So many maintainers and
contributors out there in this room are carrying so much weight and holding so much space for all
of us. I hope we can do more to help them or at the very least, I hope we can spare them kind words.
I'm not under illusions here. I don't expect what I've said here today to do all that much.
People have said a lot of what I said here many times before, but maybe, just maybe, I've touched
one heart or one mind and maybe that heart or mind will go out there and they'll make a different
choice because of what I said here today. Or maybe they'll speak up and share what was touched
today with the next person when they see something wrong. Maybe, like our very first ancestor who
looked up and reached, maybe we can make a little difference now that will make a really big difference
for the people who come after us. Because the last 10 years, the platformification of the web,
the inshidification of those platforms, that was not a new normal. That was a glance at a future
that doesn't have to be. Our power out as a community is in our principles and it's in our
numbers. If we can convene, if we can coordinate, if we can collaborate, if we can take good care
of each other and choose kindness every day, if our leaders stay humble and choose the greater
good over their own enrichment ego or fame, we can change the course of this information age.
We can change the course of history. But it will take all of us working together and it will be
damn hard work. The wonderful organizers of FOSSTEM have given me this stage, so to close this talk,
I will now issue a challenge as if all of that wasn't already a challenge. From my perspective,
and I'm speaking especially to our leaders, we must focus our collaborative energy and kindness
on the following three areas. We must make the internet more efficient. We must make our code
bases smaller. We need to reduce storage usage, duplicated requests, and reduce the distance
data needs to travel. We are in the midst of an energy and an environmental crisis.
Half our world is drowning and the other half is on fire. And as the diaspora of people across
digital social spaces continues, we must collaborate across the internet community to protect
disadvantaged, disenfranchised, and marginalized people. When diversity and inclusion suffers,
we all suffer. Our pursuit of knowledge, societal progress, and the advancement of humanity only
succeeds when we are inclusive of all walks of life, of all creeds, of all religions, of all races,
of all colors, of all communities. Barring those who promote violence or enable hate.
And we must protect science and knowledge. We must stand for the truth, not only from a geopolitical
and societal perspective, but also on an individual level. We need to protect people and the systems
through which we organize into collectives. We have to make the truth resilient. Whether you
recognize it or choose to identify as part of it, you are part of a movement. Whether you're doing
this in your spare time as a passion or as a hobby, or if you're one of the lucky people who
has found a company to pay you to do this. You are part of a movement. You have experience and
passion and you're smart as heck we need you. And I believe in us. Thank you.