►
From YouTube: Fluence, Peer-to-Peer Compute
Description
TOKEN2049 Singapore 2022, Day 1
Tom Trowbridge, Fluence Labs co-founder
A
Thank
you
very
much,
it's
great
to
be
here
at
token
2049
appreciate
you
guys
coming
we're
going
to
talk
about
compute,
which
is
the
missing
piece
in
a
web
3
world
and
I.
Think
we've
made
the
huge
advancements
across
this
web3
ecosystem,
but
we're
not
there
yet,
and
this
is
what
will
take
us
there.
So
what
is
fluence
fluence
is
a
peer-to-peer
compute
protocol
and
it
enables
this
decentralized
Computing,
which
I'm
going
to
talk
about
a
little
bit.
I
helped
found
hedera,
hashgraph
and
I'm
a
co-founder
of
fluent
slabs.
A
What
is
compute
first
of
all,
and
this
is
a
little
bit
high
level.
Most
of
you
may
understand
this,
but
it's
the
processing
that
underlies
all
transactions
that
do
not
in
all
processing,
that's
not
on
your
phone
or
your
desktop
and
that
kind
of
is
Central
to
any
type
of
complex
application.
So
where
is
compute
compute
takes
place
in
two
places
right
now,
either
the
cloud
or
proprietary
in-house
servers,
and
importantly,
the
cloud
is
not
a
benign
puffy
white
ethereal
organism.
A
It
is
dark-
and
it
is
depicted
here
as
a
rain
cloud
and
a
storm
cloud,
because
it's
dangerous
and
so
don't
buy
the
white
puffy
cloud
logos
that
you
you
see
from
some
of
the
from
some
of
the
providers
what's
wrong
with
centralized
compute.
Well,
if
it's
hosted
at
particular
companies,
it's
clear,
there's
censorship
potential
and
we've
seen
that
Facebook
Twitter
very
easy
for
a
centralized
company
to
censor
and
if
it's
at
a
cloud
the
issue
is
dependency
and
a
cloud
is
very
easy
to
get
an
application
spun
up.
A
The
other
problem
with
centralized
compute
is
de-platforming
and
we
think
of
deplatforming
in
two
different
ways.
First,
is
what
applications
can
do
and
applications
like
Twitter
Facebook
can
just
cut
you
off
for
no
reason.
We
see
it
happen
regularly,
and
that
happens
with
a
variety
of
opinions.
They
don't
like.
It
also
happens
by
the
cloud,
can
kill
applications
and
we
saw
that
happen
and
most
publicly
with
parlor
and
no
one's
life.
A
Here,
I,
don't
think
was
dramatically
impacted
by
parlors
de-platforming,
but
it
was
a
big
wake-up
call
to
people
and
companies
on
Amazon
as
the
power
they
have
for
any
violation
of
a
58-page
term
of
service
that
keeps
changing
and
keeps
growing.
And
why
does
this
happen?
Governments
governments
are
behind
de-platforming
because,
as
companies
grow,
they
have
to
interact
with
governments
and
governments
have
more
and
more
influence.
A
We
think
web3
may
solve
that.
But
if
you
get
to
web
3,
you
see
the
same
problems
and
a
lot
of
web
3
companies
actually
rely
on
the
cloud
right
now.
Infuro,
a
critical
application
relies
on
AWS
that
doesn't
it's
nice
to
have
a
decentralized
front
end.
But
if
your
computation
happens
in
the
cloud
you
can
be
turned
off
again
by
a
government
directive
by
a
term
of
service
perceived
or
real
violation,
and
that
is
not
a
comfortable
place
to
be.
A
You
also
have
web
3
dapps
that
are
not
on
the
cloud,
but
instead
run
their
own
servers.
That's
centralized
too.
They
may
have
the
best
intentions,
but
that
is
not
secure.
That
is
not
safe,
and
that
is
most
definitely
not
decentralized
and
by
the
way
this
isn't
their
fault.
It
is
just
the
fact
there
hasn't
been
an
opportunity
to
change
and
to
have
a
decentralized
compute
back
end.
So
is
this
the
internet
we
signed
up
for
I,
don't
think
that's
the
case.
A
It
was
definitely
not
Tim
berner's,
Lee's
Vision,
but
the
internet
rewards
scale
and
every
company
that
starts
to
grow.
It
gets
more
Revenue
as
it
gets
more
customers
and
you
end
up
with
oligopolies
and
even
monopolies
in
different
areas,
and
these
dominant
companies
just
will
have
relationships
with
governments.
A
That's
just
how
the
world
works
for
Better
or
For
Worse,
you
name
an
industry,
whether
it's
a
Legacy
Media
companies,
whether
it's
the
railroads
all
of
them,
had
very
deep
ties
to
existing
government
organizations
and
that
leads
to
levels
of
control
that
vary
and
we're
never
in
a
more
centralized
place
than
we
are
now
because
of
the
architecture
and
the
incentives
and
the
economics
of
the
web.
So
how
do
we
get
here?
A
We
started
off
with
mainframes
and
desktops
and
then
kicking
and
screaming
people
move
to
the
cloud
and
to
give
a
sense
of
this
type
of
transition.
I
was
investing
in
data
centers,
which
was
kind
of
the
early
description
of
the
cloud
back
in
1998.
All
right,
do
you
realize
how
early
that
was?
Aws
didn't
start
for
years
after
that
that
was
probably
10
to
20
years
too
early
to
be
looking
seriously
at
data
centers
we're
now.
The
next
generation
of
that
is
peer-to-peer
platforms.
The
cloud
is
dominant
80.
A
The
the
British
regulator
just
announced
that
81
of
cloud
revenues
in
the
UK
are
Amazon,
Google
and
Microsoft,
and
there
is
now
a
an
Anti-Trust
investigation
about
those
Cloud
companies.
That's
how
dominant
they
are,
but
that
took
25
30
years
to
get
there.
This
next
world
is
the
peer-to-peer
World
we're
at
the
very
beginning
of
it
and
what
is
peer-to-peer
compute
peer-to-peer
compute?
Is
this
compute
but
happening
in
a
decentralized
way
without
a
centralized
host?
Interestingly,
you
can
run
it
on
a
cloud.
You
can
run
it
in
a
centralized
company.
A
You
can
run
it
in
a
house
but
you're
in
an
open
platform.
People
can
switch
and
move
their
data
around
easily
and
seamlessly,
and
all
those
entities
do
is
provide
the
infrastructure
and
the
open
peer-to-peer.
Compute
is
a
protocol
that
uses
that
Hardware
infrastructure
and
allows
compute
to
happen
and
move
easily.
A
There's
two
types
of
compute
and
everyone
here
is
very
comfortable
and
can
knows
blockchain
compute,
that's
smart
contracts
and
smart
contracts
are
peer-to-peer
compute
and
they
work,
but
they're
only
sufficient
for
deterministic,
compute,
they're,
great
and
perfect
for
defy,
and
that
will
work
and
continue
to
work
forever,
but
their
deterministic,
redundant,
constrained
and
expensive
and
slow
general
purpose.
Compute
is
what
99
of
the
web
runs
on
and
that's
what
we
need
to
move
web
2
into
web
3
and
have
proper
decentralization.
A
That
is
what
fluence
is
working
on:
permissionless
with
universal
runtime,
that's
cheap
and
fast,
and
infinitely
scalable.
That's
where
we
think
that's
what
we
have
to
get
to
because
scaling
blockchain
compute
to
billions
of
transactions.
A
second
like
the
web
is
is
not
a
realistic
goal.
So
why
is
this
the
future?
Because
peer-to-peer
is
scalable,
secure
and
resilient?
A
It
can
scale
faster
and
cheaper
when
you
have
a
hardware
marketplace
around
the
world
that
can
respond
to
price
much
faster
than
any
centralized
company.
So
a
global
Marketplace
for
hardware
for
compute,
whether
it's
computer
storage,
is
infinitely
scalable
and
can
respond
much
faster
than
any
centralized
company.
A
Also,
it's
secure
it's
censorship
resistant
because
there
is
no
one
host.
There
is
no
one
company
that
either
hosts
or
provides
that
application,
and
so
there's
no
legal
entity
to
compel
no
management
to
threaten
no
directors
to
influence
and
no
shareholders
to
pressure
so
open
peer-to-peer
networks
and
Protocols
are
the
only
chance
we
have
to
resist
the
power
of
governments
dominating
and
controlling
kind
of
every
aspect
of
the
web,
which
is
the
direction
we're
heading
in.
It's
also
resilient
peer-to-peer
networks.
A
A
A
At
infrastructure-
and
so
this
is
something
also
critical
for
the
for
the
well-functioning
kind
of
world
that
we
want
to
live
in,
and
peer-to-peer
compute
platforms
right
now
exist
in
payments
right,
you're,
not
seeing
Bitcoin
shut
down
in
any
of
these
countries
and
you're,
not
seeing
you
know,
five
ipfs
or
file
coins
shut
down
in
these
countries.
They
work
and
that's
terrific,
but
we
don't
have
compute
and
without
compute.
A
You
can't
have
that
type
of
environment
that
works
for
actual
functioning,
complex
applications
and
so
look
at
the
traction
we
have
Traction
in
peer-to-peer
payments
with
Bitcoin
ethereum,
that's
not
going
away
and
we
have
Traction
in
storage.
The
storage
is
kind
of
the
great
story
of
peer-to-peer
over
the
last
several
years,
especially
with
filecoin
just
hitting
200
petabytes
of
storage.
That's
a
huge
amount
of
storage.
There
is
a
long
way
to
go,
but
they
are
scaling
and
so
that
I
think
problem
is
well
on
its
way
to
being
solved.
A
But
peer-to-peer
compute
in
terms
of
smart
contracts
is
Tiny
and
how
tiny
hard
to
gauge
this
Market?
It's
a
couple
hundred
million
dollars-
maybe
it's
300,
maybe
it's
500!
Uc
estimates.
All
over
and
it's
going
to
grow
quickly,
it's
going
to
get
to
multi-billions,
but
how?
Let's
compare
that
in
scale?
Cloud
revenues
are
126
billion
in
one
quarter,
all
right
on
a
year.
That's
about
4
000
times
more
and
Cloud
encompasses
content,
delivery
services,
Encompass
storages.
So
it's
not
just
compute.
A
So
that's
not
a
fair
Apples
to
Apples
comparison,
but
it
gives
you
orders
of
magnitude
scale
as
to
what
we're
talking
about
here
and
shows
how
tiny
smart
contracts
are
in
comparison
to
what
we
need
to
do
in
order
to
get
really
functioning
systems
going.
And
why
is
that?
It's
because
it's
hard
and
why
is
it
hard?
You
need
to
be
decentralized,
verifiable
and
scalable,
and
so
without
blockchain
that
that
that
verifiability
is
a
challenge.
A
And
so
fluence
is
using
a
combination
of
web
of
trust,
Quorum
and
ZK
proofs
to
achieve
that
security,
and
when
you
have
that
which
is
the
verifiability.
And
then
you
do
that,
you're
not
constrained
by
the
blockchain
limitations
and
the
competition
isn't
blockchain.
The
competition
is
AWS
and
the
other
Cloud
providers,
and
when
you
get
to
that
world,
which
is
where,
where
fluence
is
where
what
do
you
get?
What
do
you
enable
well
no
closed
ecosystems?
A
You
can
have
open
applications
that
are
easily
composable
and
shareable,
and
that
is
lead
to
terrific
Innovation,
because
it's
a
creativity,
accelerant
you
can
build
easily
on
top
of
other
applications.
You
can
be
sure
you're
not
cut
off
you
can
when
applications
and
code
is
hosted,
they
can
be
hosted
by
anybody
and
can't
be
turned
off
as
long
as
there
is
one
single
host
available.
So
it's
Unstoppable
and
that's
where
we
want
to
be-
is
a
very
creative,
very
secure,
Unstoppable
world,
and
what
does
that
world
look
like?
A
It
looks
like
the
most
beautiful,
transparent
Cloud
on
a
summer
day
that
you
barely
even
see
because
the
cloud
is
out
there,
but
it
is
so
light
and
so
thin
that
you
almost
don't
even
see
it.
That's
the
type
of
cloud
that
is
okay,
that's
not
a
storm
cloud,
that's
the
cloud!
You
want
where
you
and
it's
when
you
marry
storage
and
you
with
compute
and
obviously
payments
are
a
critical
part
of
that.
A
That
is
what,
where,
where
we
are
going
and
the
vision
that
I
think
we
all
will
benefit
from
achieving
and
where
is
fluent
stand
in
this
we've
launched
the
peer-to-peer
Network.
So
the
peer-to-peer
computation
network
is
live.
It
functions
the
language
Aqua,
which
allows
the
native
composition
of
peer-to-peer
applications
is
live.
It
abstracts
out
all
the
complexity
of
building
peer-to-peer
applications,
the
peer
Discovery,
the
failover
that
is
solved
with
aqua.
So
you
can
compose
applications
with
just
several
lines
of
code.
It's
a
terrific
innovation.
A
This
year
we're
launching
the
Dow
it's
going
to
be
a
Swiss
Foundation.
It
will
govern
the
fluence
protocol
and
that's
where
we
are
right
now
and
then
next
year,
we're
adding
on
the
compute
economy
and
that's
sort
of
like
the
file
coin
to
the
ipfs
and
that
will
allow
hosts
to
be
paid
and
allow
open
source
software
providers
to
share
in
that
hosting
Revenue.
A
As
the
code
they
write
is
used,
which
is
what
we
think
will
inspire
a
terrific
amount
of
code
to
be
on
this
open
peer-to-peer
world
and
lead
to
ever
growing
innovation
and
so
find
us
at
dash.fluence.dev.
You
can
find
us
at
fluence
underscore
project
on
telegram
and,
let's
think
about
what
kind
of
world
we
want
to
live
in.
So
we
want
three
companies
controlling
internet
hosting.
Do
you
want
nation
state?
A
Censoring
all
manners
of
either
applications
or
Communications
peer-to-peer
and
open
source
is
the
only
way
we
have
to
prevent
that
world
from
happening,
and
so,
let's
think
about
you
know,
our
mission
here
is
going
back
to
Tim.
Berner's
lead
about
the
goal
of
the
web
is
to
serve
Humanity.
We
don't
even
know
what
might
be
coming
when
Tim
created
the
HTTP
protocol.
Do
you
think
he
could
have
imagined
Uber,
no
chance
right,
and
so
what
we're
doing
is
trying
to
create
the
platform
and
the
rails.