►
Description
The latest innovations in decentralized storage and blockchain technology are unveiled at FIL Paris 2023. Join Molly Mackinlay, Protocol Labs's lead for engineering and research, as she delves into the Top 3 Filecoin Breakthroughs in 2023. The video highlights Filecoin's mission to create a resilient and efficient foundation for humanity's information, the Filecoin master plan, and critical updates on storing, retrieving, computing, and scaling the Filecoin ecosystem.
Molly's insight into the remarkable achievements in robust retrievals, programmable compute, and global scalability presents a roadmap for Filecoin's current and future contributions to Web3. Her insider view explores areas like the Filecoin Virtual Machine, Lassie, Saturn, and more, all aimed at achieving faster retrievals, compute over data, and a decentralized storage network.
Learn how the Filecoin network, the world's largest decentralized storage network, is powering transformative changes in data handling and security. Discover how over 1.2 exabytes of data are being onboarded and how innovative technologies like Filecoin Saturn enable web3 CDNs and fast retrievals.
Engage with the future of data and blockchain technology. Watch now and become part of the Filecoin ecosystem!
Speaker: Molly Mackinlay
#DecentralizedStorage #filecoin #filecoinfoundation #web3 #blockchain #DWeb #cryptocurrency #crypto #decentralization #Filecoin #FILParis2023 #Blockchain #MollyMackinlay #DataOnboarding #FilecoinBreakthroughs #IPFS #GlobalScalability
A
I'm
pleased
to
Welcome
to
the
stage
protocol
labs's
lead
for
engineering
and
research
Molly
McKinley,
who,
in
addition
to
be
a
being
a
world-class
engineer,
is
also
a
writer
whose
book
goodbye
to
all
of
that
is
well
worth
a
read.
Everyone
please
welcome
Molly
to
the
stage
to
tell
us
about
the
top
three
file
coin
breakthroughs
of
this
year.
Thank
you.
B
Someone
clearly
Googled
me
I,
feel
found
out
way
back
in
the
day
in
a
previous
life,
I
worked
on
the
Google
search
team
and
learned
by
working
with
Google
books
that
you
could
become
more
well
known
in
your
Google
search
results.
If
you
published
a
book,
so
I
published
a
book
and
it's
actually
not
that
hard
I
do
highly
recommend
it,
but
yeah.
B
B
Cool
all
right,
hello,
everyone,
thank
you.
So
much
it's
exciting
to
chat
with
you
all
today
would
love
to
share
a
little
bit
of
my
vantage
point
of
some
of
the
really
exciting
things
I
narrowed
it
down
to
three,
and
that
was
challenging
I,
maybe
I'm
kind
of
cheating
and
I'm
accounting.
A
couple
that
I
think
are
big
breakthroughs
of
2023
that
we've
already
done.
I'm,
not
counting
those
but
I
wanted
to
dive
into
some
of
the
really
exciting
work.
B
So
first
off
the
mission
of
filecoin.
This
should
be
really
familiar
to
all
of
our
friends
here
today
is
to
create
a
decentralized,
efficient
and
robust
foundation
for
humanities
information,
and
that's
really
by
working
with
clients
around
the
world,
helping
them
onboard
their
data
onto
filecoin,
so
that
filecoin
can
help,
maintain
it
and
store
it
long
term,
with
much
cheaper
storage
prices
with
a
much
more
resilient
foundation
for
their
information,
not
putting
it
in
centralized
data
sources
or
intermediaries
and
then
make
it
useful
to
them.
Make
it
accessible
make
it
a
foundation.
B
They
can
depend
upon
in
the
future
and
start
providing
more
useful
services
on
top
of
that
data
storage,
retrieval
compute
and
much
much
more
foreign.
This
might
also
be
familiar.
It's
the
filecoin
master
plan.
Juan
helped
sum
up
this
amazing
kind
of
ecosystem
Mission
and
organize
it
into
these
three
steps.
First
was
to
build
the
world's
largest
decentralized
storage
Network,
which
check
mark
we
have
achieved.
There
are
thousands
of
storage
providers
across
tens.
You
know
44,
plus
different
countries,
storing
exabytes
and
exabytes
of
capacity
making
that
available
to
individuals
around
the
world.
B
B
And
so
now
there's
you
know
over
150
unique,
daily
onboarders.
Those
are
storage
providers
that
are
offering
a
data
onboarding
services
over
1,
750
clients,
unique
clients
who
are
storing
data
with
you
know
a
large
portion
of
that
storing
over
a
hundred
terabytes
of
data,
and
as
I
mentioned,
we
crossed
that
six
petabyte
per
day
threshold,
which
I
will
tell
you
a
year
ago,
seemed
impossible
and
now
there's
over
1.2
exabytes
of
data
onboarded
onto
filecoin.
By
all
of
these
amazing
amazing
groups
and
I.
B
Think
that's
something
absolutely
that
we
can
celebrate
it's.
It
is
a
breakthrough
that
has
happened
over
the
course
of
last
year
and
this
year,
but
it
sets
us
a
really
good
foundation
to
then
bring
additional
Services
additional
capabilities,
additional
tooling
to
this
foundation
and
there's
actually
a
really
amazing
website.
If
you
want
to
go,
engage
with
what
are
people
storing
right
now,
you
know
how
might
I
work
with
this?
B
As
you
can
see,
these
are
large-scale
data
sets
as
well.
These
are
not
just
you
know,
your
initial
proof
of
concept
to
show
that
this
is
working.
This
is
massive.
You
know
genomics
data
sets
massive.
You
know
testimonials
videos,
you
have,
you
know
blockchain
data,
that's
getting
onboarded,
Etc
click.
B
Thank
you
and
these
really
span
a
whole
series
of
different
use
cases,
and
so
that's
one
of
the
amazing
things
about
working
with
all
of
these.
You
know
world-class
Partners
to
bring
their
data.
We
have
data
that
documents.
You
know
what's
happening
in
Ukraine
from
the
Starling
lab.
We
have
cern's
Atlas
project,
that's
storing,
neutrino
data
and,
like
let
me
tell
you
when
you're
talking
about
large
data,
you
do
not
know
large
data
until
you
until
you
start
talking
to
CERN
and
man.
B
Do
they
collect
a
lot
of
data
that
they
want
to
make
accessible
to
researchers
around
the
world
and
they
want
to
you
know
harness
so
that
we
can
find
the
next
Higgs
boson
or
whatever
the
more
recent
one
was.
So
we
can
learn
more
about
the
world
around
us
and,
like
that's
the
sort
of
data
that
we
need,
a
robust
foundation
for
humanities,
information
for.
B
All
right,
and
so
that
that's
a
a
domain
step,
two
we've
seen
immense
progress
in
the
past
year
and
a
lot
to
celebrate
there,
and
that
brings
us
on
to
step
three.
Having
1.2
exabytes
of
data
stored
in
filecoin
starts
asking
the
question:
well,
what
can
I
do
with
that
data?
How
do
I
make
that
data
more
useful?
B
How
do
I
interact
with
that
data
to
solve
problems
as
a
client
and
user,
and
that's
what
step
three
is
about
bringing
compute
to
the
data
enabling
web
scale
applications
to
interact
with
and
build
on
top
of
that
data,
making
that
data
useful
to
the
clients,
who've
stored
it,
and
that
really
boils
down
into
kind
of
four
main
parts.
We've
already
talked
about
verifiable
storage.
We've
made
immense
progress
there,
that's
a
significant
building
block
and
achievement
in
and
of
itself,
but
then
there's
three
other
really
important
areas
that
are
my
top
three
breakthroughs
for
2023.
B
That
I
think
are
still
work
in
progress.
That
I
want
to
talk
with
you
more
about
today,
which
is
robust,
retrievals,
programmable
compute
and
Global
scalability
robust
retrieval
is
talking
about
all
of
the
work
that
Falcon
Saturn
is
doing,
and
a
number
of
other
retrieval
markets
as
well
to
enable
global,
fast
access
to
data
that's
stored
across
Falcon
storage
providers.
B
There's
also
work
being
done
around
light,
retrieval
clients
and
retrieval
incentives
to
make
sure
that
when
you
put
your
data
in
filecoin,
you
can
access
it
and
load
it
through
things
like
the
ipfs
Gateway,
seamlessly
programmable
compute.
Obviously,
we've
done
immense
progress
launched
in
March
this
year
around
the
Falcon
Virtual
Machine
continues
to
be
amazing.
Work
happening
there,
but
the
really
exciting
part
is
not
just
compute
over
the
state
of
the
filecoin
blockchain.
B
It's
unlocking
compute
over
data
and
there's
awesome
progress
that
folks
here
heard
from
Ali,
but
I'll
give
you
a
quick
recap
as
well
around
how
Buckle,
Yao
and
lilypad
and
other
projects
are
really
bringing
compute
to
those
data
sets,
which
is
a
huge
opportunity
for
both
filecoin
storage
providers
and
filecoin
storage
clients
who
now
want
to
run
ml
algorithms.
Who
want
to
do
you
know
Transformations
over
their
data
to
make
them
more
useful
or
tuned
for
certain
applications.
This
is
a
really
really
exciting
opportunity.
B
What
should
we,
as
web
3,
be
aiming
for
as
an
ecosystem
so
that
we
can
have
have
all
of
the
applications
and
tools
of
web2
be
able
to
build
upon
the
same
web3
Primitives
that
we
think
should
be
The
Primitives
underpinning
the
web
at
large,
and
how
do
we
actually
help
enable
those
folks
to
come?
Join
us
on
this
Quest,
so
diving,
first
into
retrievals.
B
B
They
earn
Falcon
rewards
through
an
fevm
smart
contract
which
pays
out
and
validates
these
nodes
as
they're,
providing
their
their
retrieval
services
and
Saturn
is
currently
on
porting
the
ipfs
Gateway
at
ipfs.io
as
their
first
production
client,
which
serves
around
1.2
billion
requests
per
week.
And
this
is
you
know
the
perfect
proof
point
of
offering
very
fast
access
to
data.
B
And
you
know,
after
getting
the
ipbs
Gateway
super
happy
in
production,
we'll
be
moving
on
to
a
number
of
other
groups.
So,
if
you're
interested
to
hear
more
about
Saturn,
you
want
to
maybe
join
web3.
storage
and
metaplex
on
our
waitlist.
Come
talk
to
me.
I
would
love
to
to
know
what
you're
excited
about
loading
and
accelerating
across
ipfs
and
filecoin.
B
B
Hates
me
this
point:
we're
just
going
to
do
a
little
circle
and
we're
going
to
figure
out
where
it
might
there.
We
go
all
right
back
where
we
started
number
two
of
my
top
three
breakthroughs
for
2023
is
programmable
compute.
As
folks
know,
Falcon
Virtual
Machine
launched.
Oh
now,
it's
now.
It's
really
snazzy
launched
earlier
this
year
and
this
brought
smart
contracts
and
programmability
to
filecoin,
and
so
we've
now
seen
a
plethora
of
really
interesting
applications
start
getting
launched
from
defy
to
storage
applications
and
far
beyond.
This
is
fully
evm
compatible.
B
So
all
of
these
are
you
know,
running
in
solidity
and
fbm
was
designed
especially
to
have
a
hypervisor
layer.
So
not
only
can
we
plug
in
evm,
which
is
our
first
runtime
in
fvm,
but
we're
already
working
on
bringing
our
next
runtime
to
fvm,
so
that
more
and
more
different
developers
can
can
bring
their
applications
and
it
can
connect
Falcon
even
more
closely
into
other
ecosystems.
A
B
B
Some
of
the
work
that's
happening
across
nft
storage
and
Lighthouse,
where
they're
actually
making
recurring
filecoin
storage
deals,
you
you
know,
give
them
an
endowment
up
front
and
they
automate
the
process
of
monitoring
and
renewing
those
storage
deals
long
into
the
future,
which
I
think
is
fantastic
and
definitely
something
that
we
can
now
add
to
the
repertoire
of
all
of
those
useful
client
interfaces
in
in
filecoin.
We
have
some
really
interesting
stuff
happening
around
compute,
which
I'll
talk
about
in
just
a
second.
B
Obviously,
retrieval
is
now
utilizing
those
on-chain
payment
contracts
and
then
some
really
awesome
work
happening
around
data
Dows
and
nfts.
Helping
give
individuals
who
are
storing
data
control,
an
agency
over
the
data
that
they
store,
that
they
curate
that
they
collect
so
that
they
can
say
make
that
available
to
compute
or
ml
algorithms
being
run
on
filecoin.
B
Okay,
cool
so
I
mentioned
compute
over
data
is
the
the
the
star
of
the
show,
not
just
starting
with
actually
being
able
to
to
run
those
programmable
smart
contracts,
but
building
layers
of
compute
on
top
and
definitely
the
the
amazing
example
here
is
all
of
the
amazing
AI
work.
That's
happening
all
the
new
machine,
learning,
algorithms
that
are
getting
deployed
and
wanting
like
desperately
looking
for
access
to
any
GPU.
They
can
get
their
hands
on
in
order
to
scale
their
training
or
refine
their
algorithms
with
new
input
data.
B
And
so
this
is
a
really
exciting
opportunity,
not
just
for
filecoin
with
all
of
this
data
to
train
over.
But
with
all
of
these
storage
providers,
who
have
resources
amassed
that
they
can
now
also
put
to
work
within
compute
over
data
networks
so
that
they
can
contribute
those
and
also
receive
additional
Revenue
sources
for
the
work
that
they
do
and
so
the
exciting
that
we
expect
filecoin
to
be
used
not
just
for
one
compute
over
Data
Network,
but
for
many
compute
over
data
networks.
B
Thanks
I'll,
just
tell
you
when
I
want
to
go
to
the
next
slide,
so
Baku
Yao
is
the
ipfs
of
compute
over
data.
It's
designed
to
be
a
modular
Computing
library
that
allows
you
to
quickly
spin
up
networks
of
nodes
and
underpin
many
of
those
l2s
with
common
building
blocks
that
they
want
to
use
to
run
their
jobs,
common
job,
spec,
so
you're
able
to
Pipeline
and
interoperate
between
many
of
these
networks
and
gets
you
up
and
running.
B
What's
super
exciting
about
Lily
Pad
V1,
instead
of
v0,
which
is
a
new
test
net,
specifically
utilizing
baklao,
where
other
people
can
start
bringing
their
own
compute
nodes
and
runs
using
this
kind
of
modicum
flavored
evaluation
method,
an
ability
to
to
assemble
those
networks,
assign
compute
jobs,
smartly
across
that
Network,
so
that
you
know
hey.
If
there's
data
co-located
with
this
particular
compute
node,
we
can
intelligently
locate
the
compute
with
the
data,
so
we
don't
have
to
re-download
it
over
a
small
bandwidth
and
then
it
can
evaluate
that
those
jobs
are
being
done
correctly.
B
You
know
assign
kind
of
like
fault
fees
if,
if
anything
is
found
to
be
incorrect
and
then
accelerate
this
whole
test
net,
it's
still
very
early
days,
they've
only
just
gotten
it
set
up,
but
we're
going
to
be
welcoming
more
folks
to
bring
their
own
compute
nodes
to
that
Network
in
the
coming
weeks.
So
grab
that
QR
code.
If
you're
interested
in
learning
more
about
lilypad,
there's
a
ton
of
docs
about
where
we're
at
right
now
and
then
steps
we're
taking
to
run.
B
You
know
entirely
on
fvm
utilizing
IPC
so
that
not
only
can
Lilly
Pad
as
a
network.
You
know
operate
as
a
layer
two
on
top
of
filecoin,
but
it
can
also
recursively
Shard
and
even
have
Regional
subnets,
which
I'll
tell
you
more
about
in
just
a
second,
so
that
you
can
do
very,
very
fast
compute
jobs
in
local
local
networks
or
local
areas.
B
B
It's
just
a
test
net,
but
the
aim
is
that
we'll
be
able
to
unlock
additional
rewards
for
people
who
can
kind
of
amass
their
resources
and
offer
them
to
folks
who
are
you
know
paying
for
compute
jobs,
starting
with
both
looking
at
the
AIML
use
case,
especially
fine-tuning
things
using
Laura
and
also
the
data
onboarding
use
case?
Where
not
only
do
you
want
to
transform
data
when
it
first
comes
onto
filecoin
to
make
sure
it's
all
content
addressed
and
car
filed
and
great?
B
So
all
nice
and
easy
all
right
last
but
not
least,
is
around
Global
scalability
and
making
sure
that
we
have
kind
of
a
web
scale
foundation
for
all
the
work
that
we're
doing
here.
So
you
might
have
heard
of
IPC
or
interplanetary
consensus,
which
is
really
focusing
on.
How
do
we
move
hundreds
of
transactions
per
second?
How
do
we
scale
to
reach
some
of
these?
B
You
know
YouTube
twitch,
large-scale,
fast-paced
applications
that
can't
go
through
a
tiny,
tiny,
tiny
consensus
bottleneck
and
if
we're
going
to
really
bring
those
applications
into
web3
and
have
them
be
bringing
say,
like
a
you
know,
a
web
3
game
where
every
single
interaction
is
not
just
you
know,
purchasing
some
nft
but
actually
interfacing
with
each
object
inside
that
video
game
ends
up
on
the
blockchain.
We
just
need
drastically
better
scalability
than
we
have
today,
and
so
the
this.
B
B
But
we
are
gearing
up
for
that
Milestone
by
refining
the
the
node
Runner
so
that
when
you
spin
up
these
networks,
it's
really
easy
to
get
started
and
also
auditing.
All
of
our
smart
contracts
that
exist
today
to
make
sure
that
they
are,
you
know
all
buttoned
up
and
safe,
and
so,
if
you
would
like
to
register
here,
are
our
mainnet
contracts
and
our
calibration
net
contracts
so
go
ahead.
Go
grab
an
alpha
taste
of
what
IPC
can
bring.
B
We're
really
excited
about
this,
not
just
for
the
compute
over
data
use
case,
but
we're
definitely
excited
about
this
for
Lily
Pad
as
a
test
net,
but
also
for
many
others
who
want
to
bring
different
capabilities
to
their
applications
and
there's
a
couple
of
things.
We're
really
looking
for
with
a
good
web3
scalability
framework.
First,
we
want
to
Target
internet
scale
applications.
We
want.
B
I've
ever
heard
of
of
is
being
able
to
unlock
ephemeral
or
subsequent
subnets,
which
can
run
very
quickly
by
being
regionally
bounded,
say
you
want
to
spin
up
a
subnet
just
within
West
Coast
or
just
within
a
specific
data
center.
That
sort
of
recursive,
sharding
or
recursive
scalability
is
something
that
IPC
unlocks
it's
aiming
to
make
it
super
lightweight
to
spin
up
each
one
of
those
different
subnets
and
then
re-merge
State
through
checkpoints
with
the
parent
chain,
and
so
this
kind
of,
like
you
know,
dynamic
flow
of
hey
I
need
more
capacity.
B
Okay,
cool
I'm,
good,
now,
I'm
going
to
re-merge
State
between
networks.
That's
the
sort
of
thing
that
can
help
us
attack
local
demand.
It
can
help
us
be
much
more
efficient
with
you
know,
say
doing
a
whole
ton
of
data
onboarding
in
a
specific
data
center
or
with
a
specific
client
and
that's
the
sort
of
foundation
that
we
want
for
all
of
the
applications
we're
building
and
last
but
not
least,
we
also
want
a
a
blockchain
runtime
time
that
we're
able
to
upgrade
much
more
easily.
B
How
wonderful
would
that
be
if
anyone's
developed
a
an
L1
blockchain?
You
know
how
amazing
that
sounds.
That's
the
vision,
and
so
those
are
the
three
breakthroughs
that
I
think
are
super
super
exciting
all
within
this
theme
of
kind
of
compute,
over
data
and
enabling
web
scale
apps,
but
bringing
really
robust,
retrievals
helping,
add
programmability
and
compute
onto
filecoin
data
and
then
scaling
globally,
all
right,
and
so,
if
anyone
wants,
wants
a
screenshot
but
otherwise
I'm
happy.
If
we
have
time
for
questions
I'm
happy
to
take
any
otherwise.