►
From YouTube: Voice Gems: A 1000 Year Archive - Harry Yeff
Description
In 2020 Harry Yeff and Trung Bao set out to preserve a set of human voices for 1000 years, a feat that evokes many challenges and questions. In this talk & performance, artist and technologist Harry Yeff (Reeps100) showcases his portfolio of speed-of-light radio transmission, AI synthetic voices and how emerging technologies provide an artistically explosive and vital discussion on new methods of preservation.
A
A
The
most
important
part
of
Art
and
experience
in
this
context
is
separate
to
technical
challenges.
The
Zeitgeist
perception
of
a
specific
technology
has
huge
huge
influence
on
the
retention
and
the
accessibility
of
said
technology.
So
not
all
artists
are
relevant
to
that
challenge,
but
when
we
think
about
well
what
is
this
new
breed
of
creative
technologist
that
can
actually
work
with
engineers
and
in
some
way
have
a
deeper
understanding
of
what
is
success?
A
What
is
the
actual
use
cases
for
the
technology
and
my
most
recent
collaboration
is
actually
curating
the
artistic
intelligence
series,
which
is
a
collaboration
with
itu
and
United
Nations,
highlighting
specific
artists
that
can
help
deepen
our
understanding
of
Technology.
So
the
most
important
problem
that
we
have
are
narratives
of
fear.
The
sentiment
is,
is
that
if
you
can
make
knowledge,
explosive
and
experiential,
this
does
have
some
practical
contribution.
Art
is
nice.
Beauty
is
great,
but
the
idea
is
like.
Can
you
transfer
the
narrative?
A
A
But
what
I
found
Most
Fascinating
is
many
of
the
radio
frequencies
are
still
traveling
at
the
speed
of
light
out
into
space,
and
it
made
me
think
about
well
actually
the
media
that
we
create
the
content
that
we
make
it's
a
very
difficult
challenge:
the
concept
of
preservation,
but
most
important.
There
is
this
concept
of
time
and
I
think
that
the
narratives
that
we
have
the
way
that
we
think
about
preservation
is
actually
needs
a
lot
of
work
for
the
general
Zeitgeist.
A
So
what
are
some
other
interesting
examples
of
how
we
can
approach
that
the
second
contextualization
is
like?
Why
are
voices
important?
So
why
am
I
spending
my
life,
focusing
specifically
on
that
forgetting
about
the
music
is
that
language
on
average
has
about
30
phonemes.
There
are
languages
that
have
more,
but
if
you
compare
those
pieces
of
language
to
the
actual
expression
of
voice
you're,
looking
at
about
10x
on
average
people
use
about
20
of
their
vocal
range
and
I'm,
particularly
interested
in
the
non-linguistic,
the
phenomena
of
voice.
A
So
if
you
think
holistically
on
Earth,
whether
it's
artificial
or
human,
to
human
connection,
the
human
voice
is
an
absolute
fundamental
part
of
our
relationships
of
our
loved
ones.
Our
working
relationships,
but
also
like
non-language,
is
something
that
needs
to
be
highlighted
if
I
spent
my
whole
life
speaking
just
like
this,
how
would
it
affect
my
relationships,
my
loved
ones?
It
would
change
the
quality
of
my
life,
so
range
and
experimentation
and
vocal
phenomena
is
an
absolute,
absolutely
fundamental
part
of
day-to-day.
A
A
A
So
that
very
very
strange
recording
is
a
generative
version
of
my
voice
using
sample,
RNN
and
I.
Think
one
new
narrative,
that's
becoming
more
and
more
common
in
terms
of
preservation
and
Legacy
is
the
context,
the
content
concept
of
the
second
self.
So
these
vast
amounts
of
data
that
we're
producing
more
and
more
opportunities
are
arising
to
be
able
to
interact
with
that
data.
Obviously,
artists
are
exploring
the
concept
of
the
second
self
and
AI
second
self
spawning.
A
There
is
new
language
coming
around
this,
but
there's
more
and
more
of
a
responsibility
for
people
to
actually
see
this
as
a
very
fundamental
part
of
their
Mirror
Image
in
life.
So
how
is
that
being
kept?
Where
is
that
being
kept
and
I?
Think
it's
really
interesting
to
think
how
we
are
preserving
our
Legacy
using
Technologies
like
this?
A
A
second
element
of
my
experimentation
is
switching
modality,
so
voice
is
very
much
like
smoke.
It's
this
ephemeral
thing
that
shifts
and
changes
and
collecting
thousands
upon
thousands
of
sounds
around
the
world.
I
realized
that
there
were
opportunities
to
visualize
the
human
voice,
so
I've
been
designing
generative
systems
for
around
10
years.
We've
generated
hundreds
of
vocal
sculptures
and
I'm
going
to
show
you
an
example
here
and
I'm
going
to
share
a
little
bit
of
context.
A
So
this
is
a
project
called
C
Sound,
where
we
open
it
up
to
the
public,
where
hundreds
of
individuals
could
come
in
and
visualize
their
voice,
and
the
importance
of
that
is
when
you
change
the
modality
of
a
piece
of
media.
Something
very
special
can
happen.
The
meaningfulness
of
experiential-
because
this
is
the
key
point
I'm
trying
to
make
leaving
media
in
its
linear
form,
is
incredible,
but
to
open
up
a
new
way
to
experience
that
Media
or
that
technology
can
have
some
interesting
effects.
A
A
I
think
is
a
very,
very
important
thing,
so
the
most
recent
project
is
voice
gyms,
the
1000
year
archive
and
I'm
actually,
here
to
really
ask
you
all
a
question
which
I'll
get
to
the
end,
but
this
is
the
most
recent
generated
system
that
I've
designed,
which
is
voice
centered
and
effectively.
It
uses
the
voice
to
generate
a
high
resolution.
A
A
Whether
the
same
recording
is
used
twice,
you
arrive
at
the
exact
same
work,
so
those
fingerprint
like
features
place.
The
200,
000
particles
and
features
like
resonance
and
articulation
is
actually
results
in
the
specific
coloring.
So
it's
a
very
simple
data
visualization,
but
what's
actually
happening
with
this
project,
was
half
expected
and
half
unexpected,
so
I've
been
reaching
out
to
a
set
of
remarkable
individuals
and
the
manifesto
of
voice
gems
is
to
preserve
the
world's
most
unique
remarkable,
but
also
vulnerable
voices.
A
Here
we
have
sugwen
Chung
an
operator,
suguen
who's,
an
incredible
AI
artist
had
a
dialogue
with
one
of
her
second
selves,
her
machine-based
relationships,
an
operator
did
something
interesting.
They
actually
sent
in
a
poem
which
they
both
read
at
the
same
time.
So
we
did
two
readings
off
both
of
their
voices
generated
two
gems
and
actually
produced
what
we
call
a
fusion
voice
gym
outside
of
just
literal
contribution
to
voice
voice.
A
Gems
is
an
opportunity
to
holistically
explore
vocal
phenomena,
and
this
was
the
first
set
of
voice
gems
that
we
actually
placed
on
chain
called
the
internet
gems.
We
collected
2
000
voices
from
Anonymous
individuals
on
the
internet.
They
submitted
these
bizarre
interactions,
some
meaningful
some
playful,
some
strange,
and
we
use
that
to
generate
a
set
of
20
20
Works,
highlighting
that
Anonymous
anonymity
that
happens
online,
so
we're
also
interested
in
vocal
phenomena.
More
importantly,
institutions
started
to
reach
out
to
us
so
an
incredible
individual
called
Ben
mirin
who's.
A
A
So,
preservation
isn't
just
a
a
human
need,
but
the
ability
to
think
laterally
and
generate
these
objects
and,
through
the
set
of
rules,
start
to
develop
this
visual
language.
We
now
have
120
pieces
and
we're
aiming
for
thousands
and
having
having
these
that
sheer
amount
of
objects.
Each
with
a
very
unique
story
is
quite
a
challenge
like
where
does
that
live
permanently?
A
The
sentiment
is
a
thousand
years,
but
what's
the
actual
technical
possibility
with
that
so
immediately
when
we
started
to
share
that
challenge
more
this
idea
of
digital
ceremony
and
that's
a
very
un
Tech
conference
event
kind
of
term.
But
what
is
the
spiritual
and
ceremonial
potential
of
decentralized
preservation
like
what
are
the
stories
and
the
opportunities
for
emerging
Tech
projects
to
tap
into
that
deeply
emotional
utility
that
many
Traditions
attempt
to
do
so?
A
Something
that
we
never
expected
is
we
started
to
receive
hundreds
and
hundreds
of
requests
of
people
that
wanted
to
preserve
what
we
see
as
small,
sometimes
not
precious
media,
but
if
you
have
recordings
or
moments
on
your
phone,
what
we're
finding
there's
a
huge
demand
for
new
tradition.
So
this
is
a
very
unique
piece.
It's
one
of
my
favorites
and
I
just
want
you
to
listen
to
it.
First
tell
me
your
name:
what's
your
name:
where
did
you
go
today.
A
Hey,
hey
such
a
simple
small
moment:
children's
voices
always
produce
brighter
gems
and
we
isolated
her
voice
and
removed
the
fathers
just
for
the
rendering
and
the
the
generative
elements.
But
this
piece
is
a
dialogue
between
a
father
and
a
daughter.
It's
something
that
would
exist
as
a
voice
note
or
on
cold
mechanical
devices,
and
when
these
moments
pass
on
whether
it's
a
loved
one
that
isn't
here
anymore
or
A,
Moment
Like
This,
where
a
child
grows
up,
there
is
more
and
more
a
need
to
explore.
A
A
A
The
mummification
of
concept,
spoken
word
speaking,
is
the
first
decentralized
system
all
the
way
through
to
deep
chain
opportunities.
So
I'm
here
to
ask:
how
would
you
preserve
a
voice
for
a
thousand
years?
There
will
be
more
and
more
artists
like
me
that
have
these
full
spectrum
portfolios
and
approach
these
Technologies
as
experiential
knowledge
and
attempting
to
be
explosive.
But
how
can
we
find
the
most
incredible
stories
and
pair
it
with
the
best
examples
of
technology?