►
From YouTube: [Swarm Mini Summit] Towards fair data society
Description
Gregor Žavcer, Tadej Fius, Daniel Nickless
We live in a world where individuals are exploited for their personal data and privacy as human right is ignored while handful of actors are unfairly amassing wealth generated by personal data. We are the product, lacking any rights. In short, we can argue that we live in times of increasing data slavery.
Fair Data Society is a vision and initivate that seeks to provide an alternative. Initiated by Datafund project, it follows a vision where humans are at the center, have full control over their digital selves while the value of data is – contrary to popular centralised services – fairly distributed. Data exchange is on a need-to-know basis, according to user consents and conditions while human rights and privacy are respected. To achieve this, FDS is developing fair data protocol where personal data is semantically organised and stored on Swarm, and Fairdrop dapp for ethical data exchange. FDS components can be reused by any other project. Moreover, FDS is actively connecting various stakeholders such as developers, academia, regulators, NGOs and governments to drive social change and raise awareness towards mentioned goals.
The talk will briefly present FDS initiative and give a demo of the Fairdrop dapp and PoC demo of our compute node, concluded with a call-to-action to build projects in an ethical way on fair data principles.
Bio/company pitch:
Datafund project guards personal data, provides safe storage and enables provable personal data exchange. Information is shared anonymized or on a need-to-know basis according to an agreement between exchanging parties. Datafund’s mission is to be an enabler of a fair data society.
A
Hello
from
data
fault-
we're
happy
to
be
here
and
thank
you
all
for
coming.
Every
time
it's
won
some.
It
feels
like
a
little
family
gathering.
So,
let's
start
we
have
a
lot
of
things
to
show
today,
so
we'll
be
just
kind
of
blazing
through
it.
The
title
of
the
talk
is
towards
fair
data
society.
So
what
do
we
mean
with
by
that?
It's
without
further
ado,
if
I
say
Facebook
is
not
fair
data
or
the
Amazon
or
Google's,
and
so
forth.
I
think
the
intuition
is
already
quite
clear,
so
fair
data
in
general.
A
We
came
up
with
that
concept
and
we
believe
that
the
digital
world
needs
sort
of
like
an
equivalent.
We
see
in
the
fair
trade
movement
and
with
fair
data.
We
illustrate
how
the
new
data
economy,
which
respects
human
rights
and
privacy,
and
thus
fair
distribution
of
value,
should
should
work
so
brief
overview.
That's
data
fault.
We
were
all
in
on
swarm
so
and
we
build
on
top
of
it.
The
nonprofit
part
we
call
fair
data
Society
and
we
wish
that
it
has
like
it
become,
gets
like
life
on
its
own.
A
So
please
also
consider
this
as
an
invite
to
join
and
within
that
we
have
a
bunch
of
stuff.
We
have
the
front
end.
The
fear
of
death
denote
the
protocol
itself
and
lots.
We
plan
lots
of
developer
support
with
libraries
to
make
it
easy
for
developers
to
build
new
kind
of
applications
on
fair
data
principles.
Then
the
protocol
itself
is
neutral
and
on
top
of
it,
there's
actually
the
data
front
platform
which
should
provide
kinda,
the
regulative
environment
and
services.
A
So
the
idea
in
general
is
we
have
a
big
vision,
but
we
believe
we
should
start
small.
Think
big.
Also
when
we
look
at
the
data
situation
currently
in
the
world,
it's
like
we
could
easily
just
you
know,
lock
ourselves
and
work
develop
for
three
years,
but
we
really
want
to
provide
as
soon
as
possible
some
utility
value
and
also
through
that
in
this
cycle
of
providing
solutions
also
that
should
feed
the
protocol
development
in
itself.
So
what
are
we
going
to
see
today?
B
Okay,
so
I'm
willing
a
person
sees
they
fared
rock
which
have
been
working
over
the
last
few
months.
It's
basically
an
aim
at
doing
a
we
transfer
clone,
which
works
in
a
decentralized
environment.
So
then
this
gives
you
some
idea
of
kind
of
the
restrictions
that
are
put
in
place
to
get
the
paid
app
like
the
encryption.
You
only
get
password
symmetric
encryption
on
the
paid
version
and
everything
else
is
unencrypted
and
there's
25
million
users,
which
so
I'm
told,
is
actually
no
more
42.
So
you
could
decide
to
do
like
this,
but
totally
decentralized.
B
This
gives
you
an
idea
of
kind
of
where
we've
gone,
so
we
transfer
you're,
probably
familiar
with
it,
emailed
links
no
encryption
or
symmetric
encryption
paid
all
on
the
web
to
cloud-based
platforms
pretty
insecure.
So
we
decided
to
go
instead
with
more
a
theory
and
based
approach,
asymmetric
encryption,
elliptic
curve,
and
then
we
use
that
in
s
subdomains
for
user
names
and
ended
up
with
mailboxes,
so
that
literally
all
you
need
is
a
client
and
any
aetherium
node,
that's
running
swarm
so
put
the
team
together,
it
was
I'm
being
app
development.
B
B
A
Right,
thank
you,
then.
So,
maybe
a
little
bit
of
summary,
what
we
just
showed
now
we
basically
send
a
file
over
swore
that
was
encrypted
and
in
the
demo
you
haven't
seen
any
tedium
addresses.
There
was
just
four,
but
it
done
didn't
see
a
swarm
hash
at
the
end.
The
idea
is,
we
have
a
so
called
grandma
test
so
because
also
our
experience
shows
it's
like
decentralization.
Privacy
are
not
strong
enough
value
propositions
for
people
just
to
change
the
paradigms.
A
We
need
to
provide
the
same
level
of
user
experience
and
then
add
this
to
the
top
and
the
whole
idea
also
why
we
started
with
fair
drop
was
in
essence,
you
know
sending
file
from
person
a
to
person.
B
can
be
considered
the
basic
data
exchange,
so
we
plan
to
deepen
the
features
and
also
then
to
join
with
what
we're
going
to
demo
now
very
shortly,
a
fan
load.
So
what
is
fair
note
today
will
tell
more,
but
essentially,
users
can
run
algorithms
and
they're
on
their
data.
Users
can
be
humans,
machines
or
smart
contracts.
A
One
of
the
ideas
is
that
to
the
data
that
we
find
sensitive,
you
know.
Sometimes
it's
like.
We
don't
want
to
share
the
data
with
somebody,
because
we
are
afraid
what
will
happen
to
data
afterwards,
so
fan
Lord
brings
the
CPU
to
the
data
on
the
other
side.
Also,
sometimes
you
know
you
just
have
like
this
awesome
predictive
model
that
you
also
don't
want
to
share,
and
in
that
sense
well,
basically,
we
bring
the
CPU
day
there.
Moreover,
our
what
what
we
believe
is
pher
data
needs
new
business
models.
A
Today,
surveillance
and
dopamine
economy
are
just
not
the
right
solution
so
with
enabling
users
to
have
their
running
their
own
nodes
or
network
participants.
We
believe
we
create
up
possibilities
for
new
business
models
where
actually
the
data
can
be
treated
accordingly
to
who
generates
and
owns
it,
not
how
it's
today
oops.
So
what
we're
going
to
see
our
protocol
has
three
layers.
The
first
layer
is
constant
management.
Constant
is
like
the
permission
it's
actually
it
starts.
Always
there,
then
the
next
layer
is
personal
data
exchange.
This
we
actually
want
to
disincentivize
in
monetary
aspects.
A
It's
about
purposeful
exchange
on
the
need
to
know
basis
like,
for
example,
if
I
would
want
to
send
my
medical
records
to
my
doctor.
I
want
just
him
to
see
it
now
that
it
goes
to
some
other
people's
computers
and
the
third
layer
is
anonymize
data
exchange.
Now
also,
as
I
said
before,
we
want
to
provide
value
immediately,
so
the
solutions.
Today
we
are
presenting
solutions
for
the
first
two
layers.
It's
not
here
we're
not
speaking
about
full
anonymization
and
things
like
that.
A
Like
secure,
multi-party,
computation,
homolytic
encryption
that
this
comes
here,
so
what
we
show
in
the
proof
of
concept,
smart
contract
triggers
an
event
that
are
not
catches.
It
pulls
code
from
swarm,
pulls
data
from
sperm
compiles.
The
code
performance
calculations
saves
the
results
to
swarm
and
notifies
the
smart
contract.
This
is
now
the
first
cycle
that
we
did
so
today.
A
So
well,
today
is
setting
up
I'll,
give
a
small
use
case
example.
So
we
can
better
kind
of
picture,
it
imagine
imagine
Alice
Alice,
for
example,
she
she's
kind
of
fan
of
DNA
and
stuff
like
that,
but
she
really
doesn't
want
that
guys
like
ancestry.com,
which
take
your
DNA
forever.
Did
they
hold
it?
You
know,
so
she
has
her
own
copy
locally
locally
stored.
A
Then
she
contacts
her
doctor
for
some
can
be
recommendations,
treatment
whatever
and
basically
the
hospital
has
algorithms
now
Alice
can
actually
run
those
algorithms
on
her
data
provide
the
hospital
with
the
results
and
the
her
DNA
never
left
her
home
computer.
But
this
is
just
one
of
the
examples
it
can
go.
Also.
The
other
way,
for
example,
Alice,
sends
her
electronic
health
records
to
her
doctor
in
the
hospital,
because
the
hospital
is
big
and
has
like
good
predictive
models
and
they
did
some
tests.
They
combined
the
data
and
they
returned
the
results.
C
C
Basically,
there
are
route
hashes
right.
There
are
the
locations
yeah
and
they
have
that
content
type
and
this
can
be
can
get
complicated,
because
what
we
trying
to
do
is
like
match
the
inputs
and
outputs.
So
we
define
the
input
schema
that
goes
into
the
into
the
compute
node
and
the
output
that
the
output,
some
algorithm,
produces,
or
something
yeah
so
working
on
data
structures
all
the
way,
and
then
we
have
like
consent.
C
So
these
are
basically
compute
nodes
that
are
registered
and
we
have
like
signers
and
stuff
who
sign
it,
and
we
are
still
missing
the
proofs
of
all
this
stuff.
But
again,
this
is
a.
We
go
into
the
swarm
that
the
location
of
the
swarm.
We
pull
out
the
data
structure
where
everything
of
all
of
these
things
are
written
and
we
put
it
out
so
and
then
we
can
like
manually,
emit
jobs
or.
C
We
can
have
like
a
child
and
we
can
see
all
the
options
or,
let's
say,
algorithms,
that
you
can
run
only
five.
So
so,
if
I
check,
if
I
see,
for
example,
this
one
and
we
just
so,
we
have
like
it's
like
costs,
it
needs
to
be
executed.
We
know
where
this
one
swarm.
We
we
know
like
slacker
again.
File
is
actually
a
project.
It
gives
us
deformation
how
we
should
compile
the
source
code
into
the
assembly
and
where
the
entry
points
are
and
what
kind
of
data
do
we
want
to
put
into
this.
C
C
C
Okay,
so
we
don't
do
any
like
trying
to
do
any
kind
of
computation
validation
for
now,
it's
very
simple:
we
check
the
hash
of
the
code.
We
see
what
code
it
goes
down.
The
note
checks
these
two.
So
what
is
different
in
the
in
the
game
file
in
the
project
file
again
the
project
file.
Actually
all
the
files
are
in
the
swarm,
so
it
like
combines
all
the
stuff
into
the
the
the
project.
C
C
A
So
the
idea
here
is
we
basically
again,
this
is
the
back
end
to
what
you
saw
with
fell
drop.
It's
not
just
to
run
the
operations
like
city
operations.
This
number
20
42
is
just
running
a
script
and
not
using
any
data,
but
we
can
actually
put
in
all
kinds
of
data
and
predictive
models
now.
What
does
this
mean?
This
means
like
like
I,
said
before
we
are
with
this.
A
D
A
D
A
So
I'm,
not
the
coder,
but
I
will
try
to
give
like
the
best
possible
answer.
So
basically,
what
we
use
diffie-hellman
is.
We
use
ENS
sort
of
like
MIT
s,
public
key
directory.
So
when
we
create
a
mailbox,
we
save
the
public
key
of
that
mailbox.
Mailbox
is
basically
a
wallet
generated
locally
and
then
when
basically,
what
the
only
thing
that
users
need
is.
You
know
if
I
want
to
say,
send
the
file
to
you.
A
I
ask
you:
what's
your
address,
you
say
send
to
Amir
and
basically
we
look
up
your
public
key
and
calculate
the
difficult
one
with
between
my
mailbox
and
your
mailbox.
So
to
answer
your
question,
it's
basically
you're
going
to
see
only
what's
happening
between
us,
not
what
I
would
send
for
to
today
or
to
Victor,
and
for
this
we
are
actually
also
using
working
this
swamp.
It's
for
the
notification
part.