►
From YouTube: Advancing Human Rights with Web3: Starling Lab
Description
If fake news has you worried, its result—fake history—is way more serious. Starling Lab, jointly developed by the USC Shoah Foundation and Stanford University’s Department of Electrical Engineering, is creating a framework for combating misinformation, tracing provenance and reducing information uncertainty across a variety of applications, from news media to historical preservation.
A
So
it
feels
frankly
like
a
decade,
and
so
my
goal
today
here
is
to
cover
with
you
all
of
the
incredible
things
that
we
have
done
at
the
starling
lab
and
collaboration
with
many
friends
that
are
here
at
this
talk,
and
then
I
want
to
paint
for
you
a
vision
of
the
the
future
as
to
where
we're
headed
and
some
of
the
table
stakes
about.
A
Now
here,
there
was
a
web
3
movement
that
was
starting
to
create
a
new
type
of
conversation
about
how
we
could
use
advanced
forms
of
cryptography
to
reshape
the
internet.
And
what
dawned
on
me
was
that
this
opportunity,
at
a
technical
level,
also
had
an
amazing
opportunity
to
affect
what
we
think
of
as
civics
or
the
idea
of
civil
society
on
the
internet.
And
so
what
we
thought
over
at
stanford
and
at
usc
was
that
we
could
engage
early
in
this
discussion
to
think
about
the
role
of
data
integrity
and
what
we
meant
by.
A
A
So,
rather
than
moving
fast
and
breaking
things
and
trying
to
really
drive
adoption
as
the
first
priority,
our
goal
was
actually
completely
different.
It
was
really
the
inverse
which
was
to
think
about
social
impact
from
the
beginning
and
to
establish
a
set
of
principles
that
could
guide
the
ethical
and
the
regenerative
approach
to
creating
this
type
of
technology,
as
opposed
to
an
extractive
and
breathless
approach.
A
So
we
began
by
looking
at
the
ways
in
which
we
could
use
data
integrity
to
help
different
domains,
and
so
this
the
first
place
we
started
was
with
history,
and
so
we
chose
a
vulnerable
data
set,
which
is
the
vote.
The
data
set
of
the
usc's
visual
history
archive,
which
comprises
55
000
survivors
of
genocide's
testimony.
You
can
see
here
on
the
right
that
the
genocide
collections
span,
the
early
part
of
the
20th
century
with
the
armenian
genocide
all
the
way
up
to
contemporary
crises
in
myanmar,
with
mass
violence
against
the
rwinga.
A
We
have
focused
on
creating
tape,
storage
solutions
in
the
past,
with
the
usc
show
foundation,
and
so
what
we
needed
to
figure
out
was
a
way
of
downloading
the
information
from
tape,
packaging,
putting
it
into
offline
files
like
car
files
and
then
going
and
automating
deal
making
in
an
at
scale
at
a
four
petabyte
scale.
So
I
have
a
lot
of
people
to
thank
in
the
incredible
work
that
we've
done
over
this
year.
A
It
was
literally
dozens
of
people
have
come
forward
and
then
also
the
mining
community
has
been
in
or
the
storage
provider
community
has
been
really
important
in
helping
us
scale
from
a
phase
one
of
81
tabi
bytes
to
the
the
four
pebby
bytes
of
arcful
quality
information,
and
so
we're
engaged
right
now
in
the
process
of
phase
one
and
quickly
moving
to
phase
three.
A
So
for
folks
that
are
here
and
and
interested
in
working
with
us,
please
do
get
in
touch,
because
we
have
quite
a
bit
of
information
that
we
are
going
to
spread
and
we
want
to
decentralize
this
as
much
as
possible
to
different
organizations
all
around
the
world.
A
Then,
as
different
storage
opportunities
go
up
and
down
for
malicious
or
potentially
reasonable
concerns
around
and
failures,
then
of
course
we
need
to
think
about
repair
strategies
that
will
be
able
to
automatically
ensure
that
we
have
enough
replications
of
the
data
in
the
ecosystem
and
then
finally,
there
is
a
really
important
thing
to
consider
in
terms
of
liquidity,
as
we
are
scaling
and
trying
to
put
many
more
petabytes
of
data
on
the
network,
we
want
to
make
sure
that
storage
providers
of
all
sizes
are
able
to
put
up
the
pledges
that
they
need
in
order
to
store
this
information
and
that
will
require
them
seeking
new
forms
of
capital
and
assistance
to
get
that
done.
A
Let
me
switch
gears
to
the
journalism
program,
so
this
here
was
an
incredible
journey
that
we
started
in
march
with
the
2020
election.
So,
just
before
the
pandemic
shut
everything
down,
we
were
able
to
go
out
with
reuters,
photographers
and
think
about
ways
in
which
we
could
establish
a
chain
of
custody
in
the
cameras
themselves
and
then
move
the
proof
of
the
hash.
Essentially
that's
taken
at
source
up
to
the
decentralized
cloud,
and
so
there
you
can
see
in
some
of
these
demos
here
that
we
deployed
in
the
field.
A
We
were
using
professional
grade
cameras
and
then
cryptographically
enabled
mobile
phones
to
really
leverage
things
like
secure
enclaves
that
could
do
advanced
hardware,
signing
of
hashes
and
then
from
there
it's
pretty
incredible.
We
were
able
to
move
things
into
the
actual
cms
of
the
reuters
newswire,
and
you
can
see
here
in
this
particular
example.
From
march
we
were
able
to
take
that
file,
and
then
you
can
see
here
there
are
our
hashes
that
were
established
at
source
that
are
now
actually
stored
in
the
reuters
cms.
A
So
it's
pretty
exciting
to
then
take
all
that
information
and
start
to
put
it
onto
the
decentralized
web
and
we
working
with
adobe.
One
of
the
things
we
were
able
to
do
was
go
to
the
next
level
of
decentralization,
so
here
we
took
files
and
actually
injected
into
the
metadata
of
the
file
itself,
a
universe
of
information
so
that
you
could
see
here,
like
let's
say,
you're
on
a
website.
You
could
click
on
this.
A
I
and
then
see
information
around
who
took
this
photograph
when
they
took
it
where
they
took
it
and
then
also
the
links
back
to
the
decentralized
web.
So
you
see
here
falcon
cids,
ipfscids
et
cetera,
so
that
process
was
really
inspiring
because
it
took
us
then
into
the
fall,
and
we
worked
with
reuters
to
deploy
this
technology
at
scale
to
create
an
archive
called
the
78
days
which
catalogued
the
photos
from
every
day
from
the
election
all
the
way
up
to
the
inauguration.
A
Now,
certainly,
there
was
no
way
we
could
have
anticipated
what
happened
in
those
crucial
78
days.
But
I
can
tell
you
how
amazing
it
was
to
sit
there
and
at
the
end
of
this
process
at
the
inauguration
and
look
back
and
think
about
all
the
incredible
photographs
that
were
taken
and
now
sealed
into
a
historical
record
that
we
can
now
very
clearly
provide
a
chain
of
custody
for,
and
that
is,
it
was
a
first
for
journalism.
A
So,
as
there
are
various
misinformation
campaigns,
we
can
ensure
that
the
information
is
valid
and
and
remains
intact.
So
that
allows
us
to
do
many
things.
We
can
protect
witnesses.
We
can
help
ensure
that
there's
proper
context
around
imagery
and
ultimately
can
be
used
by
prosecutors
to
seek
justice.
A
But
what
we
found
was
that
actually
the
most
important
thing
was
that
we
needed
to
be
proactive
about
addressing
the
potential
manipulation
of
imagery
as
it
is
taken
from
the
field,
all
the
way
up
to
the
various
platforms.
And
so
we
identified
many
ways
in
which
the
data
was
vulnerable
and
we
identified
this
gap
between
the
raw
footage
that
exists
in
the
field
and
also
expert
witnesses
that
sit
in
trial
and
we've
been
pioneering
new
forms
of
what
we're
calling
supervisory
testimony.
A
They
can
leverage
advanced,
metadata
and
cryptography,
but
then
actually
have
a
human
sit
there
in
the
court
and
provide
extensive
information
about
how
the
best
practices
were
deployed
with
the
collection
of
this
evidence,
so
that
in
and
of
itself
is
showing
you
really.
The
importance
of
what
we're
doing
here
is
that
we're
really
helping
translate
advanced
technologies
into
something
that
can
be
very
practical
and
useful.
A
A
We
can
think
about
our
best
practices
and
institutionalizing
them
through
courses
and
various
publications
and
finally,
looking
at
directly
engaging
with
a
more
systematic
approach
with
our
in-field
work
by
providing
long-term
fellowships
for
folks-
and
this
is
what
ultimately
has
resulted
in
the
creation
of
the
starling
lab,
which
is
supported
by
both
pl,
as
well
as
the
falcon
foundation
for
the
decentralized
web.
We
are
so
proud
to
be
the
very
first
academic
research
institution
that
is
fully
focused
on
web
3
and
the
advancement
of
human
rights.
Think
about
that.
A
It's
taken
a
little
while,
but
now
we
have
two
major
research
universities
that
are
supporting
this
critical
effort,
and
this
very
first
and
that
will
allow
us
to
help
us
think
about
rebuilding
trust
in
all
the
domains
that
I've
spoken
about
to
you
here
today
in
history,
law
and
journalism.
We
have
extensive
programs,
we're
bringing
faculty
from
across
the
world
to
engage
around
these
issues,
and
so
I'll
conclude
here
now
with
thinking
about
what
is
this
opportunity.
A
Last
year
I
came
as
I
was
talking
about
a
lot
of
the
prototypes
that
we
were
just
starting
to
work
on.
I
talked
about
the
importance
of
decentralization
and
gave
an
analogy
and
thinking
about
how
decentralization
is
really
like
diversity
or
biodiversity,
which
is
that
the
more
types
of
diverse
the
more
diversity
you
have
within
a
system,
ultimately
the
more
resilient
it
is
and
that
at
a
technical
level
makes
sense
biologically.
A
All
of
our
amazing
hopes
and
our
excitement
about
this
opportunity,
and
we
have
now
a
label
and
we
have
a
whole
community
and
we
have
a
technology
movement,
that's
become
incredibly
powerful.
It's
become
profitable
and
now
really
is
this
this
moment
to
reshape
the
internet.
A
A
A
The
consequences
that
we
had
around
web
2
formed
because
we
just
put
too
much
pressure
on
web
2
to
solve
everything
technically
politically,
and
so
let
me
give
you
some
consequences
and
some
or
some
sorry
learnings,
that
we've
had
just
along
the
way
so
think
about
the
mobile
phone
technology
that
we've
created
and
how
exciting
it
was
for
us
to
find
ways
of
taking
metadata
to
help
establish
time
and
place
of
a
photograph.
Well,
our
researchers
have
found
that
actually,
this
exact
type
of
information
has
already
been
sorry.
A
A
A
And
yet
I
can
trace
it
and
follow
the
logic
around
how
the
most
ardent
defenders
of
bitcoin
have
been
able
to
make
the
case
about
how
this
is
now
a
way
of
pushing
back
against
all
sorts
of
economic
imperialism
that
has
existed
for
centuries.
We
can
trace,
for
instance,
like
the
good
neighbor
policy
from
the
30s
into
the
brenton
woods
to
things
like
the
currency
crisis
that
existed
in
argentina.
A
That
case
can
be
made,
and
I
understand
that
at
the
same
time,
el
salvador
is
by
no
means
a
beacon
of
human
rights
and
if
the
example
of
in
february
of
2020,
in
which
the
president
marched
troops
into
the
legislature,
to
get
some
of
his
policies
in
forest,
I
don't
think
that
that
very
much
blows
well
for
where
we
might
be
headed
with
enabling
an
economy,
and
maybe
a
government
structure
with
a
crypto
utility.
It
just
means
that
we
need
to
be
very,
very
careful
that
the
euphoria
again
is
the
danger.
A
And
finally,
we
thought
about
tokens
as
a
way
of
being
able
to
cross
borders
and
create
this
new
transnational
movement.
And
yet,
if
we
look
and
we
compare
to
the
hatching
rates-
let's
say
in
china,
with
cryptocurrencies
from
2019
to
2021-
we're
seeing
that
china
was
able
to
successfully
eliminate
all
hashing
of
bitcoin
and
crypto
mining.
A
And
so
when
we
thought
about
these
movements
like
such
and
web
2
that
we're
able
to
take
down
dictators.
Well,
we
learned
really
the
hard
way
that
putting
this
idea
of-
let's
say
facebook
as
saving
humanity
was
absolutely
the
wrong
approach.
And
so
central
bank
currencies
or
in
some
respects
in
a
similar
way,
could
pose
similar
challenges,
especially
in
a
place
like
china,
with
the
social
credit
system
that
could
be
matched
with
a
central
rate
currency
to
produce
some
pretty
challenging
results.