►
From YouTube: Tool House Rock #10 | TogetherCrew
Description
MakerDAO Forum Post: https://forum.makerdao.com/t/tool-house-rock-10-togethercrew-tuesday-june-20th-2023-14-00-utc/21123
Description:
Tool House Rock is a structured community call to give projects a way to showcase their tools and solutions to the growing MakerDAO Ecosystem. Tool House Rock will follow a consistent call agenda, focused on giving Ecosystem Actors a foundation for reviewing and assessing tools.
This Session's Guest: Danielo (daniel.ospinabotero@gmail.com)
What is TogetherCrew?
TogetherCrew, Is a data analytics tool that leverages the large digital footprint of online communities and scientific research to generate relevant and actionable insights.
TogetherCrew is designed for:
*Assessing the underlying causes driving community health
*Mapping and analyzing community interactions and behavior
*Building communities using a calculated and measurable process
Disclaimer
This communication is provided for information purposes only. The views and opinions expressed in this meeting are those of people involved and do not reflect the official policy or position of MakerDAO or any of its contributors, Core Units, or affiliates. This communication makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only and should not be relied upon as legal, business, investment, financial, or tax advice. You should consult your own advisers as to those matters. References to any digital assets and the use of finance-related terminology are for illustrative purposes only and do not constitute any recommendation for any action or an offer to provide investment, financial, or other advisory services.
MakerDAO Links:
https://makerendgame.com/toolshed
https://makerendgame.com
https://mips.makerdao.com
https://forum.makerdao.com
https://chat.makerdao.com
A
A
A
Real
quick,
some
disclaimers,
our
views
are
our
own
and
are
not
representative
of
maker
Dao
as
a
whole.
Inaccuracies
are
possible
information
changes
very
quickly
in
this
space,
especially
if
you're
viewing
this
video
in
the
distant
future.
And
finally,
none
of
this
is
financial
or
legal
advice,
purely
informational.
A
B
There
we
go
so
we
are
together
crew.
We
help
communities
come
and
stay
together.
So
first,
a
quick
overview
of
where
we're
coming
from
we're.
Seeing
these
technological
Evolutions,
starting
from
broadcasting
like
radio
television
Etc
around
the
1950s
and
the
most
effective
way
to
reach
the
people,
is
just
pay
for
an
advert
there
and
get
it
to
as
many
people
as
possible,
and
so
we
have
essentially
the
key
metrics
at
the
time
is
kind
of
like
how
many
people
can
we
get
this
in
front
of
then,
with
the
the
internet,
social
media
Etc.
B
We
have
these
rides
around
the
2010s
of
the
pay-per-click
and
targeted
advertising,
so
we
can
send
to
a
specific
demographic
and
we
pay
for
how
many
people
are
clicking
on
the
ads.
But
as
we
get
now
a
AI
open
protocols,
Etc
defensibility,
starts
to
become
a
very
different
game
like
we
saw
the
the
rise
of
the
the
content
marketing,
but
also
now
we
are
saturated
with
content
everywhere.
So
content
itself
doesn't
mean
much.
What
starts
to
have
a
meaning
is
humans.
Relationships
is
community
in
a
world
that
is
global
and
you
can
go
anywhere.
B
It
Feels
So
daunting
that
we
have
like
the
opposite
reflect.
We
want
to
go
to
smaller
groups,
where
we
can
trust
people
where
we
can
cut
through
the
noise
and
so
on,
and
have
a
essentially
a
small
trusted
community
that
we
can
really
believe
in,
and
so
we
start
to
have
also
businesses
following
this
trends
and
catching
up
with
what
these
fundamental
human
structure.
That
is,
that
we
live
in
communities
and
have
been
doing
that
through
thousands
of
years
so
for
businesses.
This
is
a
matter
and
I
mean
here.
B
Businesses
in
a
very
general
sense,
like
I'm,
also
talking
about
Dao's
protocols,
public
goods
Etc,
but
like
what
we're
talking
about
here
in
terms
of
how
they
can
relate
with
their
their
users
or
customers,
is
areas
like
loyalty,
meaning
retention
and
helping
these
people
have
a
sense
of
belonging.
So
they
want
to
stay
around
is
being
able
to
could
cut
through
the
noise
through
peer
recommendations,
especially
more
and
more
important,
as
we
have
ai,
creating
all
sorts
of
kind
of
dream
worlds
and
targeted
advertising
and
so
on.
B
What
we
want
to
know
is:
who
can
we
trust
and
we
can
trust
people
who
are
a
little
bit
like
us?
We
can
trust
their
recommendations,
so
the
concept
of
advocacy
becomes
increasingly
important
in
communities
and,
of
course,
we
have
value
creation
from
feedback
ideas,
suggestions
in
the
case
of
downs,
actually
contribution
participation
in
governance
Etc.
B
However,
the
issue
is
that,
throughout
all
of
these
building,
Community
requires
a
tremendous
amount
of
work
like
putting
a
tremendous
amount
of
emotional
work
to
build
relationships,
and
so,
when
we're
trying
to
strategize
around
it
when
we're
trying
to
figure
out
what
is
it
that
we
should
focus
on
in
community?
Are
we
doing
a
good
job
or
not?
Is
this
giving
us
the
the
expected
results?
B
Do
we
have
a
strong
community
that,
when
the
shock
happens,
we'll
join
shoulders
joint
shoulder
to
shoulder,
or
on
the
other
hand,
is
it
going
to
fall
apart,
and
how
can
we
tell
the
difference
and-
and
so
the
key
problem
that
we
have
is
that
fundamentally
Community
Builders
are
shooting
in
the
dark.
We
don't
have
a
way
to
assess
these.
We
don't
really
know
where
we
should
focus
to
have
the
most
impact
in
terms
of
building
a
community.
B
So
as
a
result
of
that,
we
end
up
doing
a
little
bit
of
spray
and
pray
doing
all
kind
of
initiatives
following
best
practices,
but
these
best
practices
are
not
even
proven,
so
we
end
up
wasting
a
lot
of
time,
money
and
overall
missing
on
a
bunch
of
growth,
and
so
what
we're
doing,
if
we
start
to
look
into
it,
is
like
okay.
So
what
are
we
paying
attention
to
and
as
we're
still
very
much
stuck
in
metrics
that
come
from
kind
of
like
the
1950s
or
2010?
B
So
we
have
the
passive
audience
metrics
of
like
how
many
people
are
watching.
So
how
many
followers
do
you
have?
How
many
members
do
you
have
in
this
board
and
so
on?
But
this
gives
us
very
little
meaning
these
people
might
be
engaged
or
not,
they
might
be
active
or
not,
they
might
care
about
it
or
not.
We
don't
really
know,
and
then
we
have
the
engagement
metrics
like
retweets
and
likes,
but
they're
very
Flicker,
and
so
the
issue
with
engagement
is
that
it's
relatively
easy
to
measure
and
from
the
whole
web
too.
B
B
Lot
of
tension
in
the
community
and
people
are
unsure
and
they're
fighting
each
other
and
name
pulling
each
other
engagement
metrics
go
up,
and
then
they
go
down
pretty
suddenly.
Equally,
we
have
some
communities
that
are
relatively
quiet
and
everyone
tremendously
cares
about
them
and
others
that
are
very
noisy
and
they're.
Just
a
thing
of
the
moment,
and
in
two
months
they
will
have
had
disappeared.
B
Also
as
well.
The
more
the
engagement
metrics
start
to
go
up
the
more
we.
We
have
risks
of
burnout,
because
people
are
hyper
engaged
they're
participating
too
much.
That
doesn't
necessarily
mean
it's
the
most
sustainable
nor
the
most
effective
form
of
Engagement,
because
we
have
a
lot
of
a
lot
of
noise
happening.
A
lot
of
people
sharing
information,
but
is
from
all
of
these
engagement.
B
Can
we
actually
translate
these
to
something
productive?
Are
we
actually
understanding
each
other?
Is
it
actually
really
working
to
satisfy
our
needs?
The
things
that
brought
us
into
the
community
and
so
on,
and
so
what
we're
fundamentally
arguing
here
is
that
even
the
community
is
becoming
increasingly
important.
What
we
have
been
measuring
is
not
really
community
and
because
we
measure
the
wrong
things,
we
pay
attention
to
the
wrong
things.
We
focus
on
the
wrong
things
and
we
get
counterproductive
outcomes.
B
So
we
started
about
a
year
ago,
seeing
Community
managers
asking
how
do
we
really
measure
Community?
We
know
that
likes
and
retweets
are
not
enough.
So
how
do
we
really
measure
it
and
what's
come
out
of
that?
First,
a
research
project
and
now
developing
a
product
to
bring
these
to
people
is
together.
B
True
and
fundamentally,
what
we're
trying
to
do
is
enable
to
people
to
have
precise
and
effective
Community
Building
through
insights,
personalized
recommendations
and
then
using
all
of
these
insights
and
recommendations
to
as
well
bring
in
some
automations
that
can
make
Community
Builders
more
effective.
So
we
try
to
eliminate
the
guesswork
and
using
the
analytics.
So
you
can
have
the
actionable
insights
and
recommendations
and
then
replace
the
tedious
tasks
using
the
precise
data-driven
automations
and
through
these
increase
at
least
10x
the
impact
that
a
community
Builder
can
have
as
well
as
them.
B
Knowing
where
are
we
right
now,
so
they
can
focus
in
between
all
of
the
different
in
areas
that
they
might
need
to
focus
is
how
much
attention
should
go
to
community
right
now
versus
X
or
Ys
or
set.
And
how
do
we
do
this?
So
we
have
developed
our
own
analysis
framework,
we're
combining
on
chain
and
off-chain
data,
mostly
about
Gathering
data,
from
as
many
platforms
as
possible
for
our
community
is
hanging
out,
because
we
have
most
digital
communities
or
even
hybrid
companies,
that
we
have
a
tremendously
Rich
data
footprint.
B
But
this
one
is
spread
across
platforms,
so
you'll
have
like
the
Discord
analytics,
the
Twitter
analytics
and
so
on,
but
where
the
data
really
comes
to
life
and
we
can
really
start
to
understand
it.
A
lot
better
is
when
we
can
bring
this
data
together
across
platforms
and
don't
have
it
siled
in
the
different
places,
and
then
we
apply
primarily
two
techniques,
so
social
network
analysis.
This
is
understanding.
How
are
the
people
interacting
with
each
other
in
the
networks
and
not
just
only
what's
the
behavior
of
an
individual?
B
But
what's
the
context
of
the
behavior
of
that
individual,
in
which
position
they
are
who
they
engage
with
and
and
so
on,
and
we
also
apply
some
machine
learning
models,
text,
analysis
and
so
on,
and
with
these
we
can
start
to
move
now
from
what
is
a
passive
audience
and
move
towards
more
measuring
the
interdependence
between
community
members,
how
together
they
create
value
how
together
they
engage
each
other.
They
they
work
together.
B
They
collaborate
and
also
these
as
a
then
as
a
social
network
as
a
structure
is
it
resilient
or
not,
and
what
can
happen
to
it
under
different
conditions,
so
furthering
with
the
research
we
started
to
map
communities
as
a
series
of
nested
systems.
So
it's
understanding
that
the
community
is
not
just
a
thing,
but
a
little
bit
like
a
body
that
has
different
organs
within
the
community
also
has
different
subsystems,
different
nested
systems
that
all
need
to
be
healthy,
so
to
understand
Health
as
a
holistic
concept.
B
We
we
know
that
we
need
to
have
healthy
individuals
who
are
happy
with
it.
If
the
individual
systems
are
unhealthy,
the
community
is
going
to
be
unhealthy.
If
the
people
are
burning
out,
your
community
is
going
to
be
fragile.
Then
we
have
to
have
good
relationship
between
these
individuals
same
thing.
If
people
are
arguing,
if
they're
hating
each
other,
the
community
is
going
to
be
fragile
and
then
we
have
the
clicks
team
subgroups
that
are
forming
and
these
themselves
need
to
be
functional.
And
finally,
the
community
needs
to
be
nested
within
a
bigger
ecosystem.
B
B
B
So
if
you
empathize,
you
understand
what
the
how
their
behavior
is
evolving
over
time
and
then
you
can
help
them
chart
their
own
course
through
the
community
and
then
for
the
knowledge
management
management
side.
We
take
all
of
these
different
data
pipelines,
and
then
we
can
gather
all
of
the
all
of
the
data
from
the
community
to
also
create
an
AI
that
then
can
help
you
answer
questions
of
like
what
is
that
team
doing?
B
On
the
other
hand,
we
have
the
networks
more
towards
the
ride
that
are
more
decentralized
or
more
distributed,
and
we
have
more
many-to-many
Communications,
so
that
has
the
advantage
of
increased
resilience
less
of
a
single
point
of
failure.
But
with
these
networks
on
the
right,
the
challenge
sometimes
can
be
to
converge
and
move
forward
in
a
single
Direction.
So
alignment
starts
to
become
a
challenge
and
you
have
a
sort
of
golden
ratio
in
the
middle
that
we
can
calculate
calculate
mathematically
where
information
is
Flowing
effectively
across
these
clusters.
People
talk
enough
to
each
other.
B
This
is
how
you
are
in
this
specific
metric
and
based
on
that
issue,
a
series
of
recommendations
of
how
you
can
try
to
achieve
a
balance,
or
at
least
be
aware
of
what
are
the
dangers
and
the
negatives
that
can
happen
from
being
in
that
position
and
so
take
action.
And
now
we've
been
running
this
for
a
bit.
We
are
very
happy
to
have
some
leading
names
as
early
adopters
of
our
tool.
B
We
we
receive
some
grants
from
Ave
near
polygon,
Aragon,
cello
Etc
that
have
supported
us
in
the
journey
and
that
have
then
been
some
of
the
early
beneficiaries
of
of
our
technology.
When
we're
doing
custom
reports,
then
slowly,
we've
been
automated,
creating
those
custom
reports
into
a
live
dashboard,
and
this
is
what
I'll
show
you
then,
in
a
second
in
the
in
the
live
Dem.
B
So
let
me
took
a
video
and
let
me
put
this
for
everyday
activities,
so
you
need
to
get
help.
Here
is
the
the
main
thing
we
have
three
sections,
so
we
have
the
community
inside
section
where
you'll
have
more
tactical
day-to-day
things
like
Community
Health,
that
has
the
kpi
sort
of
Assessments
and
then
a
survey
section
that
I'll
explain
in
a
sec.
B
Now
these
as
I
was
telling
you
is
calculate
based
on
these
Network
graphs,
so
we're
not
just
calculating
engagement,
essentially
not
just
like
whether
someone
is
posting
or
something
like
that,
but
we're
calculating
interactions.
When
someone
engages
with
another
member
and
is
through
these
interactions
that
we
calculate
the
tools
so
compared
to
other
analytic
tools,
you
might
find
that
they're
less
engagement,
essentially
that's
calculated
in
the
network,
because
a
lot
of
Engagement
sometimes
is
empty.
Is
our
messages
are
spam?
B
Are
things
that
people
don't
find
Value,
so
they
don't
engage
with
it
and
for
us
what's
important
is
when
people
are
coming
in
contact
with
each
other.
So
that's
what
we're
really
calculating
here
and
then
you
can
use
these
to
figure
out
which
members
are
fulfilling,
which
role.
So,
for
example,
the
big
dots
they
are
at
risk
of
being
overwhelmed
because
they're
receiving
really
a
lot
of
messages.
Well,
the
ones
that
are
in
the
periphery,
they
might
even
disengage.
So
we
use
this
data
and
then
we
pocket
it
into
the
different
categories
here.
B
So
total
active
members,
those
who
are
new
those
were
consistently
active
and
carrying
some
organizational
memory
and
then
the
vital
members.
Then,
from
these
we
can
break
it
down
into
line
graphs
analyze
as
well
onboarding
how
people
are
joining
the
network,
whether
they're,
activating
whether
they're,
still
participating
and
taking
part,
and
we
can
see
the
member
activity
over
time
and
the
evolution
of
these
graphics
and
also
for
the
disengagement.
B
We
can
see
not
only
whether
someone
disengaged,
but
what
role
they
were
fulfilling
in
The
Social
Network
before
so
they
were
a
newly
active
member
or
they
were
consistently
active,
for
example,
or
they
were
really
a
vital
member
who
was
core
of
the
social
tissue
and
then
we're
going
to
take
very
different
actions
to
engage
with
them
or
ask
them
for
feedback
or
figure
out.
Why
they're,
leaving
or
understand
how?
We
should
react
to
this
because
say
a
new,
a
newly
active
member.
B
Maybe
they
were
just
asking
a
question:
there
is
no
need
for
them
to
stay
around
if
a
vital
member
leaves
is
very
different
and
then,
if
we
run
a
campaign
to
re-engage
members,
we
can
also
track
these
down,
seeing
how
many
of
them
have
returned
to
the
network
and
of
course
we
see
that
over
time
as
well.
So
in
the
community
health
section
I
was
telling
you
about
before
the
decentralization
score
that's
below,
and
here
above
we
have
fragmentation,
which
is
another
one
of
the
key
scores
that
we
calculate.
That's.
B
Essentially,
how
much
is
this
community
divided
into
sub
communities
that
don't
necessarily
talk
to
each
other
or
on
the
other
hand,
is
it
too
and
mesh?
Is
it
like
a
hair
bow
people?
Everyone
is
talking
with
everyone
to
the
point
that
you
can
no
longer
distinguish
what
are
the
subgroups
or
teams
and
if
it's
to
enmesh
people
are
likely
going
to
be
a
little
bit
overwhelmed,
trying
to
keep
up
with
too
many
conversations
too
many
relationships,
but
if
it's
too
fragmented
it's
going
to
be
siled.
B
So
once
we
can
calculate
it,
these
are
the
recommendations
that
we're
telling
you
and
is
based
essentially
on.
What's
your
position
specifically
of
your
community
across
this
metric
and
what
are
the
actions
that
you
can
take
to
take
it
more
to
a
balanced
place?
And
then
here
we
have
the
same
thing
like
if
it's
too
concentrated
or
too
decentralized,
which
I
was
explaining
you
before
kind
of
like
similar,
similar
thing
and
you'll,
have
different
pros
and
cons,
different
things
to
pay
attention
depending
on
your
community.
B
But
as
we
go
through
more
and
more
of
these
indicators
and
we're
building
a
few
more
working
with
different
social
and
network
scientists,
then
we
can
really
help
you
water,
the
sort
of
initiatives
that
you
should
really
pay
attention
to
to
do
this
then,
to
enrich
all
of
these
data.
We
have
the
the
pulse
survey
section,
so
this
for
now
is
a
Discord
bot
that
is
in
the
process
of
being
built,
but
instead
of
running
a
big
survey
every
few
months.
This
allows
you
to
run
micro
surveys
on
recurring
basis.
B
So,
every
two
weeks
you
can
ask
two
free
questions
to
your
members
and
then
gauge
over
time,
how
that's
evolving
and
as
well
Benchmark
with
others.
So
we
have
created
working
as
well
with
social
scientists
and
work
scientists.
Different
templates
like
this
one
here
that
you
can
use
with
predefined
questions.
This
allows
us
to
Benchmark
and
helps
us
go
a
lot
deeper
into
some
of
the
specific
cases
or,
let's
say
sub
areas
that
allow
you
to
create
a
healthy
Community
or
not
so
to
create
them.
You
just
kind
of
go
there.
B
You
can
Define
when
it
starts
when
it
stops
the
repetition,
the
channels
in
which
you
wanted
to
see
which
sort
of
members
will
get
access.
You
can
tailor
the
questions
and
so
on,
and
once
you're
done
with
it,
you
can
simply
create
the
survey
and
it
will
get
automated
as
a
recurring
whole
survey.
So
that's
the
the
core
of
the
that's
the
core
of
the
tool.
I,
hope
you
find
it
interesting
as
I
was
saying,
we're
together
crew
trying
to
help
communities
to
stay
and
remain.
B
They
haven't
actively
communicated
with
another
Community
member
and,
unfortunately,
the
the
logging
on
is
not
something
that
we
can
track
very
well.
So
there
is
a
blind
spot
around
lurkers,
but
in
the
the
research
that
we
have
seen
around
lurkers
lurkers
are
valuable
still
for
a
community
in
the
sense
that
he
can
pass
information
to
other
places,
but
their
level
of
understanding
of
the
community
is
always
very,
very
superficial.
Because.
Unless
you
grapple
with
a
topic,
your
understanding
is
always
going
to
be
a
little
bit
shallow.
B
You
might
have
heard
the
thing
that
people
say
that
if
you
read
about
the
topic,
you
understand
10,
but
if
you
try
to
teach
it
to
someone
else,
you
understand
80
more
or
something
else
like
that
is.
When
you
engage
in
a
conversation
with
other
people,
you
start
to
go
a
little
a
lot
deeper
and
that's
when
you
really
build
those
sticky
relationships
is
for
connecting
to
other
humans.
So
we're
not
counting
the
lurkers
as
part
of
the
let's
say,
engaged
community
members.
A
B
Yeah
sure
so
the
the
key
there
is
having
access
to
the
to
the
right
information
and
platform.
So
it's
mostly
about
getting
it
to
have
access
to
the
platforms
that
the
specific
Community
is
using
and
then
the
bot
itself
can
start
to
understand.
What's
the
relevance
of
that
content,
but
we
can
also
working
with
the
community,
we
can
also
give
it
give
them
some
nudges.
So
we
can
say,
for
example,
is
like
this
is
some
content
that
has
never
been
reviewed.
B
It's
kind
of
all
content,
Etc
versus
this
is
the
the
last
version
that
was
carefully
tailored
and
loads
of
people
work
on
it
together.
So
you
can
pass
some
of
those
insights
to
the
bot,
so
it
might
prefer
certain
data
sources
instead
of
others.
That
being
said,
the
way
the
the
bot
works
and
is
multiple
AI
agents
that
collaborate
together
to
create
the
response.
B
In
the
front,
you
only
see
a
single
interface,
but
in
the
back
we
actually
have
multiple
AI
agents
and
the
ways
the
structure
is
that
it
breaks
down
the
question
into
multiple
sub
questions
that
are
simpler
and
then
you
will
go
and
find
the
best
data
source
for
each
of
the
sub
questions.
Then
it
will
kind
of
pit
them
against
each
other,
create
a
summary
and
then
give
you
the
answer.
So
is
a
bot
that
is
designed.
B
Most
likely,
yes,
I'll,
have
to
check
with
my
team
how
the
logs
are
being
stored
because
I'm
not
I'm,
not
on
the
data
science
side,
so
I
don't
know
exactly
how
they
visualize
that
information.
So
I
don't
have
access
to
it
right
now,
but
I
could
come
back
and
probably
there
is
a
log
in
the
database
because
I
know
part
of
the
bot.
C
Oh
thanks,
Juan
yeah,
I,
I
kind
of
joined,
I,
guess
Midway,
so
I'm!
Sorry,
if
you
know
my
question
would
make
you
repeat
some
information,
but
I'm
just
really
curious.
I
think
you
introduced.
You
said
it's
using
machine
learning
and
do
I
understand
correctly
that
it's
using
that
to
also
measure
the
quality
of
the
discourse.
So
does
it
use
some
kind
of
sentiment
analysis
and
how
does
that
actually
work
in
terms
of
how?
How
is
the
model
trained
to
perceive
the
the
information
that
people
are
writing.
B
Yeah
sure
so,
for
now
and
and
bear
in
mind
like
this,
is
still
evolving,
but
they
we
started
first,
primarily
focusing
on
the
social
network
analysis
just
because
there
is
already
a
lot
of
of
research
on
how
different
communication
patterns
in
a
social
network
have
been
validated
to
show
the
certain
outcomes
for
an
organization
or
for
a
community.
So
we
started
by
leveraging
a
lot
of
that.
Then.
On
top
of
that,
we
are
starting
to
map
the
the
sentiment
for
the
specific
relationships,
and
this
is
still
a
little
bit
of
work
in
progress.
B
But
the
sentiment
analysis.
Essentially
the
bot
can
sorry
the
the
sentiment
analysis
tool
can
understand
whether
an
interaction
in
between
people
was
emotionally
positive,
negative.
That's
the
basic
then
building
on
top
of
that,
whether
it
led
to
a
resolution
or
not,
and
if
it
didn't
led
to
a
resolution,
then
you
can
as
well
start
to
track
in
The,
Social
Network
how
these
tension
or
conflict
in
a
relationship
is
affecting
the
network
over
time.
B
So
the
the
exact
impact
that
these
can
have
or
let's
say,
be
able
to
predict
that
if
you
have
this
amount
of
conflict
or
exactly
this
structure,
your
token
price
or
your
customer
acquisition
cost
or
the
the
number
of
new
people
joining
will
change
by
X
percent.
That's
the
machine.
Learning
algorithm
can
start
to
do
a
lot
of
that
hard
work
or
predict,
for
example,
which
members
are
likely
to
disengage
before
they
disengage.
B
B
To
advocate
for
the
changes
that
they
would
want
to
make
the
community
safer
or
equally
when
we're
detecting
conflict,
we
could
move
it
and
alert
a
conflict
resolution
mechanism,
whether
that's
an
individual
or
some
other
tool
that
can
then
come
and
repair
their
relationships
before
the
problem.
Fasters.
C
Okay,
great
just
the
follow-up
question
as
well,
so
this
does
the
model
actually
direct
us
or
point
us
towards
the
specific
pattern
that
it
noticed.
Or
does
it
just
give
a
like
a
general
evaluation
of
the
trends
and
the
conversation
and,
let's
say,
for
example,
I'm
saying
this
because
I've
done
some
like
sentimental
analysis
years
ago,
I'm,
not
the
deaf
person
but
I.
Remember
what
we
stumbled
upon
with
social
media.
B
B
However,
when
it
starts
to
become
really
interesting
is
more
when
you
combine
it
with
the
relationship
because,
as
you're
saying
like,
maybe
you
and
I
we're
we
joke
and
we're
being
a
little
bit
sarcastic
and
the
model
might
pick
it
as
negative.
But
if
we
continue
talking,
then
that
will
turn
positive
at
some
point,
probably
we're
going
to
say:
oh
yeah,
how
you
been
and
so
on,
and
that
is
a
positive
interaction.
So
you
can
more
measure
the
rate
of
change
in
relationships
rather
than
overall
positive
sentiment
or
negative
sentiment.
B
So
you
can,
you
can
start
to
get
a
feel
for
where
your
community
is
and
it's
when
you
see
the
Delta
like
the
change
in
the
indicators,
especially
if
there
are
spikes
or
or
Falls
that
you're
like.
Oh
something
is
happening,
and
now
you
have
the
context
of
how
your
community
has
behaved
over
time
to
at
that
moment,
understand
what
that
really
means
and
take
action
or
not.
D
B
So
it's
is
expanding
at
the
moment
we
started
with
Discord.
That's
already
live,
Twitter
is
coming
live
soon.
Probably
telegram
is
the
next
one.
We're
writing
a
grant
proposal
to
get
some
extra
funds,
and
and
with
that,
make
the
discourse
integration,
but
depending
it
depends
on
the
communities
that
we're
working
with
and
the
ones
the
platforms
that
they
prioritize.
That's
the
ones
we're
gonna
build
first
and.
D
B
D
Okay
and
then
the
other
thing,
how
do
you
deal
with
the
I?
Don't
know
if
there's
a
term
for
it
like
a
smart
word
like
permeability,
I
guess,
like
I,
don't
know
if
someone
is
in
several
communities
or
or
different
like
how
do
you
know
if
a
Twitter
account,
that's
interacting
with
maker,
for
example,
belongs
to
maker
or
just
the
fact
that
they
are
interacting,
makes
them
kind
of
like
a
Community
member?
Is
that
the
kind
of
like
the
premise.
B
That
being
said,
the
the
focus
in
general
is
a
little
bit
broader,
because
what
we're
seeing
is
that
organizations
are
moving
from
being
very
close
systems
where
you
can
very
black
and
white
Define,
whose
part
and
who's
not
to
something
that
is
more
continuous.
And
so
you
start
to
have
a
lot
of
people
who
participate
in
multiple
communities
and
at
the
end,
if
they
are
well
connected
to
multiple
other
members
in
your
community.
B
What
your
community
is
becoming
is
a
node
in
the
system
rather
than
an
isolated
entity,
so
kind
of
like
from
that
perspective.
It
doesn't
make
a
lot
of
sense
to
try
to
fully
differentiate
their
part
or
not
part
is
more,
they
are
in
our
orbits
and
how
much
they
are
a
vital
member
like
at
the
core
of
the
community
or
how
much
they
are
more
at
the
periphery
and
and
that
the
network
graph
will
calculate
it.
But
he
won't
give
you
a
a
binary
member
non-member
outcome.
B
I
mean
if
they
have
never
engaged
with
your
other
members
across
the
platforms.
You
won't
see
it
and
you
can
segment
by
data
sources.
So
let's
say,
if
they're
in
your
discourse,
a
Discord
and
very
engaged.
Well,
that's
going
to
count
a
lot
more
than
if
they're
in
Twitter
talking
with
other
people,
because
Twitter
is
a
more
open
platform.
So
you
can
also
make
a
bit
those
distinctions
but
same
thing.
The
boundaries
are
a
little
bit
fussy
and
we
hope
become
increasingly
fussier,
as
organizations
become
more
interconnected,
open
and
collaborative
nice.
D
D
D
Since
no
one's
interrupting,
but
when
you
say
something
is
too
decentralized
like
for
some
reason
that
almost
seems
like
it
will
go
against
all
the
ethos
in
web,
3
and
and
people
might
try
to
Lynch
with
pitchforks
in
the
public
hall.
So
I
understand
that
the
like
okay,
this
is
to
or
or
people
are
not
engaged
enough,
or
they
are
too
far
from
the
from
the
core
or
the
vision.
B
Yes,
yes,
absolutely
the
the
devil
is
in
the
in
the
details
from
a
social
technical
Network
perspective
like
so,
you
know
pure
technical
infrastructure
like
blockchains
and
so
on.
That's
a
different
thing,
not
what
we're
talking
about
here,
we're
talking
a
little
bit
more
about
the
human
layer,
but
also
how
they
interact
with
the
platforms
and
so
on
in
a
extremely
decentralized
social
network.
You
will
have
a
sort
of
graph
where
there
is
no
Central
node,
so
you
won't
even
have
notes.
B
It
looks
a
little
bit
more
like
a
fishnet,
very,
very
distributed
very,
very
divided
in
these
cases.
What
starts
to
happen
is
that
the
the
distance
between
community
members
increases
a
lot.
So
actually,
what
you
want
as
an
ideal
structure
for
information
to
flow
and
be
processed
effectively
by
the
network,
is
something
that's
called
a
small
world.
So
in
a
small
world
you
have
a
few
clusters
that
are
very
intimate.
B
It
was
a
very
different
feel
now
it's
kind
of
big
and
I
don't
feel
belonging
as
much
and
so
on.
So
if,
instead
of
that,
you
start
to
have
these
Pockets
that
form
you
can
still
retain
intimacy
and
you
can
do
a
lot
of
the
processing
in
these
pockets
and
what
you
want
to
do
is
then
separate
by
topic
or
in
some
form,
have
thematic
differences
across
the
different
Pockets.
So
this
happens.
B
Naturally,
in
an
organization
most
organizations
divide
by
function,
but
there
is
you,
can
divide
by
product
or
divide
by
geography
or
divide
by
whatever
you
want.
The
specific
way
to
divide
will
vary
depending
on
each
Community,
but
the
important
thing
is:
you
have
the
formation
of
these
subgroups
and
then
in
the
subgroups
themselves.
We
know
that
an
effective
communication
pattern
is
relatively
distributed,
but
if
it's
completely
distributed
involving
too
many
people,
usually
you
don't
have
enough
influence
in
certain
places
so
that
the
network
can
move
forward.
B
B
You
still
want
to
give
space
for
leaders
to
emerge
like
the
idea
is
not
to
get
to
a
leaderless
organization,
it's
more
like
a
leaderful
organization
than
where
there
are
many
many
leaders
and
more
and
more
people
can
become
leaders
and
take
prominent
positions
and
influence
others,
and
that
becomes
a
more
effective
and
resilient
system
that
one
where
you
have
no
one
who's
allowed
to
take
any
power.
It's
more
about
the
way
in
which
we
grant
that
power
and
the
way
we
keep
that
power
accountable.
B
Now,
if
it
becomes
too
centralized
it's
a
disaster
as
well,
it
becomes
fragile.
There
is
again
too
much
distance
between
community
members
like
the
information
architecture
of
the
network,
just
kind
of
collapses.
So
you
end
up
having
a
middle
and
if
we're
gonna
really
talk
about
our
Theory
I
would
argue
that
the
ideal
that
we
went
for
daos
is
not
centralized
is
not
decentralized,
is
more
polycentric,
which
is
a
few
centers
of
decision
making
that
overlap
a
little
bit
and
anyway.
B
That's
the
other
indicator,
because
if
the
subgroups
are
not
connected
with
each
other,
then
also
they
they
don't
align,
they
don't
collaborate,
and
you
end
up
with
a
lot
of
silos
and
the
community
splinters
and
goes
in
different
directions,
and
it's
only
a
matter
of
time
before
there
is
conflict.
So
it's
kind
of
like
a
balance.
D
A
B
Yeah,
of
course,
so
the
this
is
an
area
that
is
not
yet
fully
developed,
but
the
direction
it
goes
is
the
more
we
start
to
gather
data
from
different
communities.
We
can
start
to
aggregate
these
data
and
have
a
sense
of
how
healthy
the
ecosystem
is.
So
then,
if
the
health,
let's
say
the
health
indicators
of
your
specific
Community,
are
going
down
or
up,
you
can
compare
that
to
the
ecosystem
at
large
and
be
like
well.
Are
we
doing
something
different
or
are
we
just
following
the
trend
of
the
ecosystem?
B
And
this
is
part
of
the
ups
and
downs,
then
the
other
thing
that
we
can
do
is
start
to
calculate
related
to
a
previous
question.
You
start
to
calculate:
what's
the
overlap
in
members
in
between
us
and
other
communities,
and
by
that,
because
we
can
see
how
these
people
are
super
active
in
this
other
community
and
also
in
ours.
So
then,
these
two
communities
have
some
good
level
of
proximity
and
the
more
you're.
Oh,
you
have
an
overlap
with
other
communities.
B
That's
going
to
be
your
closed
ecosystem,
and
so
from
these
we
can
like
very
organically
bottom
up,
start
to
see
the
health
of
the
other
ecosystems.
Now
we
cannot
tell
you
whether
there's
going
to
be
a
technological
failure
in
one
blockchain
and
not
another,
so
that
goes
that's.
That
would
be
a
solution
for
a
different
tool,
but
at
least
in
the
at
the
social
layer.
A
B
Probably
by
the
end
of
next
week,
there
is
going
to
be
a
mini
version
that
is
fully
open,
that
anyone
from
the
website
can
just
install,
but
still,
if
you
want
to
have
access
to
some
of
the
more
robust
and
complete
indicators,
and
all
of
that
you
need
to
engage
with
us.
So
just
send
us
a
message
and
we'll
be
very
happy
to
to
help
you
integrate
and
figure
out
how
best
to
approach
it
for
your
community
perfect.