►
From YouTube: COGwg Rebooted Session 01: With HackFS Special Guests
Description
Exciting news! We're rebooting our Computer Over Data Working Group, where developers come together to share their approaches to decentralized compute and AI.
Join @DeveloperAlly & Silent Spring in chatting to HackFS winners DefiKicks, Daggle and Tentai about their compute over data projects!
A
B
C
A
C
A
B
Wow,
it's
been
fast
times.
Thank
you.
So
much
for
letting
me
join,
no
I
appreciate
it.
It's
it's
been
a
lot
of
fun
since
I
think
the
Cod
Summit
we
had
a
great
showing
there
and
no
I
just
wanted.
If
anything
just
say,
thank
you
to
you
guys
for
continuing
the
effort
in
the
space
is
my
audio
coming
across
okay,
it's.
B
B
That's
right,
yes,
thank
you,
yeah!
If
you
can
find
me
on
Twitter
at
Wes,
W
Floyd,
and
we
have
this
podcast,
the
DFI
podcast,
where
we
go
deep
on
webbreed
infrastructure
and
it's
with
a
former
protocol,
Labs
friend,
ishan
chordia,
and
we
try
to
just
cover
use
cases
and
developments
in
the
space
have
interviews
with
folks
to
hear
like
to
learn
more
about
where
people
are
investing
in
the
space
and
it's
just
a
sort
of
a
labor
of
love.
So
yeah,
please,
like
And,
subscribe
yeah.
A
Definitely
yeah,
let
us
know
what
you
think:
yeah
I
would
definitely
recommend,
checking
it
out
all
right.
Well,
thank
you!
So
much
Willis.
We
have
some
special
guests
on
today,
but
first
I
think
cardiker
I'll
hand
over
to
you
to
talk
about
what
we're
up
to
today
and
we're
sorry
to
see
you
go,
but
you're
welcome
anytime.
Just
you
know,
jump
on
stream.
Thank
you
for
joining
us
and
kind
of
you
know,
starting
this
amazing
group
for
us
to
keep
going
as
well.
We
really
appreciate
it.
Yeah.
C
C
B
Started
a
listing,
it's
called
the
web3
index.xyz
as
it
sounds
the
web3
index,
and
you
know
what's
funny-
is
like
a
lot
of
people
that
are
building
in
this
space.
Have
doubts
sometimes
that
there's
demand
you
know
they
may
not
know
all
the
different
decentralized
disciplines
that
need
the
the
technology
and
then
vice
versa,
people
that
are
building
decentralized,
social,
media
or
decentralized
whatever
the
capability
is
they
don't
know
that
this
plethora
of
of
new
technologies
exist,
I'll,
actually
type
in
the
chat
window.
Here.
B
D
B
So
vice
versa,
so
so
that
you
know,
because
there's
so
many
folks
that
join
the
computer
today
working
group
over
the
past,
you
know
a
few
months
and
so
just
kind
of
making
it
a
folks
aware
that
these
various
projects
exist.
I
think
can
hopefully
bring
some
Community
together
and
just
started
a
simple
place
there,
and
if
we
can
get
that,
then
then
try
to
you
know,
keep
it
up
to
date,
so
that
people
know
that
these
these
resources
exist.
C
C
D
B
C
So
just
in
terms
of
the
purpose
you
know
like,
so
we
just
sort
of
continuing
with
that
theme
in
terms
of
collaboration
and
we're
trying
to
really
look
into
the
you
know,
developing
shared
standards
and
solutions
and
like
also
communicating
the
latest
releases
and
but
also
trying
to
focus
on
like
peer
education
and
and
really
great
support
So.
Then
people
can
come
here
and
really
get
going
and
pretty
quickly
and
then
like
in
terms
of
the
agenda.
Today,
it's
pretty
standard.
C
So
so,
if
you
go
to
the
next
slide
Ali
and
we
keep
it
nice
and
short
and
sweet,
so
it's
going
to
be
a
40
minute
sessions.
We
have
the
speakers,
the
tech
chat
about
their
projects
and
then
we've
got
a
little
bit
of
a
panel.
So
just
asking
them
a
couple
of
questions
and
then
afterwards,
I'll
do
some
Community
updates
from
our
members.
So
there's
going
to
be
polybase
the
lily
pad
team
and
then
also
the
new
Atlantis.
A
C
That
yeah
just
quickly
in
terms
of
the
future
of
Cod
working
group.
So
you
know,
as
I
said,
we
just
have
to
quickly
the
the
sessions
and
the
community
updates,
but
like
in
the
future.
What
Ali
and
I
have
been
talking
about
is
to
have
this
really
practical
and
have
like
some
coding.
You
know
like
sessions
and
then
maybe
also
some
themes.
C
A
You
know
it's
all
about
the
community,
sorry,
whatever
you
want,
we'll
we'll,
try
and
deliver
all
right.
Thank
you!
So
much
so.
Firstly,
we
have
some
very
special
guests
here
today
we
have
Leo
stellen
from
daggle
and
also
Vincent,
also
known
as
debugging
feature
on
Twitter
here
to
chat
to
us.
Now
these
guys
D5
kicks
I.
Think
it's
a
bit
late
for
him.
He
must
have
missed
the
stream
or
misunderstood,
but
I'm
sure
he'll.
A
You
know
if
he'll
listen
later
or
turn
up
light
up,
but
these
guys
did
some
amazing
work
for
hack
FS,
so
hack,
FS
is
kind
of
file
coin,
slash
protocol,
Labs
Flagship
hackathon
every
year,
and
it
runs
over
three
weeks.
It's
called
hack
file
system,
which
is
why
it
gets
the
name
hack,
FS,
and
some
of
the
previous
winners
have
really
gone
on
to
be
kind
of
big
businesses
in
in
the
web
3
space.
A
So
things
like
hudlow
one
that
you
would
have
heard
of
it's
like
decentralized,
Zoom,
I've,
like
totally
forgotten
all
the
others,
but
I
know,
there's
like
at
least
three
or
four
really
amazing
ones.
Ballast
actually
is
a
great
one.
They
are
kind
of
a
deployer
for
games
and
things
like
that
and
they
work
on
ipfs,
but
there's
some
really
other
other
big
ones.
So
these
guys
are
definitely
joining
kind
of
an
elite
bunch
of
people
in
in
winning
best
use
case
for
bacquiao.
A
A
A
E
Yep
so
like
I'm,
a
computer
science
graduate
like
I
just
graduated
this
year
and
like
I'm,
trying
to
explore
hackathons
and
the
three
stuff,
and
that's
how
like
I
found
out
about
hackathl
and
I,
built
something
called
as
daggle
for
hackers
and
Douglas
and
platform
as
a
service.
We
could
say
that
at
least
for
now,
so
like
yeah,
you
can
deploy
all
your
workload,
jobs
like
within
two
to
three
clicks
and
it
will
be
deployed
to
the
network
yeah
yeah.
It's.
A
D
Yeah
sure
so
I
built
in
SDK
that
allow
you
to
show
your
AI
model,
so
it
is
very
similar
to
hacking
phase
spaces
which
you
can
host
your
model
and
live
people
in
the
aggravate
so
I
developed
that
python
SDK.
So,
instead
of
deploying
to
some
other
hoses
service,
you
can
put
it
up
and
deploy
to
ipfs
and
make
use
of
bakara
to
1D
computation.
D
Oh
yeah,
so
I
think
there
are
a
lot
of
use
case
of
demoing.
Your
AI
applications
from
translation
to
image,
processing
and
those
are
more
like
job
based
AI
models.
So
I
think
that
is.
That
is
the
kind
of
like
the
sweet
spot
with
using
the
Buckle
out,
because
you
can
have
that
like
long-running
job
opening
executing
on
the
blacker
Network
and
then
get
back
to
results.
A
D
Oh
yeah,
I
would
say:
I
have
a
background
in
more
like
in
devops,
because
the
longest
story
is
like
I
used
to
ban
my
own
startup
and
I
spent
so
much
time
on
like
deploying
those
like
devops
stuff
and
then
trying
to
sell
post
surface
and
then
eventually
I
discovered
a
biocon
and
buckle
out
so
I
I
realized
over.
There
is
the
like.
You
know
the
best
opportunities.
How
could
we
like
have
someone
to
to
host
that
surface?
D
For
me
instead
of
I
have
to
do
so
so
I've
been
exploring
the
idea
a
lot
in
last
year
and
then
for
this
hackathon
I
I
would
really
focus
on
the
AI
and
then
I
realized.
Okay,
most
data
scientists
they
want
to
like
show
their
like
or
like
demo,
their
AI
applications
to
others
and
yeah.
That's
half
of
there,
one
of
the
like
MVP
that
I
try.
E
A
A
D
So,
as
I
mentioned
like
how
you
how
you
could
like
having
someone
to
host
the
surface
for
you,
some
Open
Source
service
and
image
service
is
really
an
interesting
example,
because
not
only
AI,
models
like
how
we
apply
those
Builders
onto
it,
but
imagine
like
those
resize,
a
Corp
or
other
image,
processing
this
kind
of
services.
Okay,
what
if
someone
holds
hosts
that
for
me
and
what
I
have
to
do
is
just
to
incentivize
them?
Maybe
I
pay
them
in
balcon
and
there's
so
many
open
source
tools
that
are
over
available.
D
So
instead
of
having
those
Service
hosted
by
something
proprietary,
CMS
or
LS,
API
I
could
just
like
incentivize
people.
Do
they
do
that
on
a
distributed
world
and
I?
Believe
you
eventually,
it's
more
cost
effective,
even
yeah,
so
I
think
I
would
try
to
build
hentai
on
a
on
a
like.
A
whole
devops
and
this
microservices
yeah.
A
C
D
I
I
can
only
say
all
this
technology
is
that
you're
perfect,
pretty
quickly
right
so
yeah
every
every
month,
I
will
I
receive
the
update.
The
news
list
for
Baka
and
I
find
it
quite
exciting.
Yeah
there
are
new
features
to
Tesla,
so
yeah
you,
you
really
have
to
like
learn
really
fast
and
then
I
try
to
see
how
different
Technologies
play
together,
for
example,
in
this
hackathon
I
learned.
A
lot
like
I
learned
about
satin
in
are
the
like
poisonous
and
maybe
like
in
next
month.
D
D
A
A
E
E
B
E
E
A
E
A
A
E
A
E
A
I
think
what's
happening
here
is
that
ipfs
can
take
a
while
to
propagate
sometimes
to
the
rest
of
the
distributed
hash
tables.
So
sometimes,
especially
when
you're
going
through
a
HTTP
Gateway
connection,
it
can
take
a
while
for
that.
Ipfs
result
to
show
up
specifically,
but
I
believe
that
you
did
this
and
it
worked
and
also
you
could
get
the
job.
A
E
A
D
A
A
A
Not
at
all
that's
the
thing
about
questions
right,
as
you
know,
you
never
know.
Who
else
is
thinking
the
same
thing
and
is
really
glad
you
asked
so
I
think
I
think
it's
great
and
keep
asking
and
then
you
get
to
help
other
people,
because
you
know
answers
and
I've
seen
you
doing
that
as
well.
So
that's
how
a
commute
good
Community
works.
So
we
appreciate
that.
C
E
Yeah
like
right
now
like
we
are
still
in
the
beta
stage,
I
mean
not
even
in
the
beta
State.
It's
in
the
alpha
stage,
I
would
say
like
I'm,
the
only
one
who's
testing
it
like
once
we
go
product
like
live,
people
can
start
upload
and
yeah.
We
will
try
to
even
promote
our
product
so
that
people
can
upload
their
models
or
data
sets
and
yeah.
Something
like
that.
You
can
do
that
too.
C
D
E
A
A
E
E
C
A
I
do
like
that
about
stream
yard
actually
also
I
I,
don't
know
if
anyone's
out
in
the
community,
but
if
you
do
have
any
questions
feel
free
to
ask
us
and
we'll
be
happy
to
put
them
put
them.
Ask
these
guys
as
well,
so
yeah
amazing,
all
right,
oh
this
is
these-
are
fun
features
Vincent?
How
can
people
get
in
touch
with
you?
Is
it
the
same
as
Twitter
your
best
contact?
D
A
C
D
A
A
D
A
A
Amazing,
well
feel
free
to
get
in
touch
with
that.
Well,
thank
you
for
joining
us
guys
and
like
we're
looking
forward
to
seeing
where
your
projects
go.
Hopefully
we
can
keep
in
touch,
and
you
know
we're
about
to
do
project
updates
now.
So,
hopefully
we
keep
in
touch,
and
you
know
in
a
few
more
episodes
we'll
be
able
to
do
some
project
updates
from
you
guys
as
well
I'd
be
excited
to
see
where
these
go.
C
A
C
So
if
you
haven't
watched
it
yet,
but
there
was
a
computer
or
data
working
group
session
15
and
the
public-based
team
was
there,
which
was
a
really
interesting
talk
and,
as
you
saw,
Leo
was
actually
implementing
this
database,
so
decentralized
database,
and
you
can
see
here
it's
powered
by
zero
knowledge
proofs
and
it's
similar
to
like
it's.
You
know,
attributes
of
blockchain,
but
it's
much
faster
and
it's
still
privacy
preserving
as
well,
and
so
quick
shout
out
for
them.
C
C
C
C
A
C
C
A
B
E
C
A
Some
big
big
companies
sponsoring
this
as
well,
so
we've
just
sponsored
as
well.
The
backyard
Lily
bad
team
has
just
added
a
bounty
and
some
sponsorship,
but
I
noticed
like
stability.
Ai
is
going
to
be
there
and
I've
forgotten
everyone
else
down,
but
obviously
Al
gavera
never
mind
like
some
really
big
names
in
the
decentralized
AI
field
are
going
to
be
there.
So
I
think
you
know
and
like
I'm
excited
to
see
because
I
don't
think
I've
seen
a
decentralized
AI
hackathon
before
yes,
so
you
can
be.
C
E
C
So
you
know
like
one
of
the
what
I'm
quite
interested
in
is
governance
and
Alignment,
so
you
know
like
in
terms
of
harm
and
bias.
How
can
we
look
into
that
and
then
you
know,
obviously,
you've
got
personalization
with
privacy,
so
you
know
like
how
can
how
can
that
be
sort
of
Federated,
Learning,
Blocks
and
and
all
of
that
yeah,
it's
very
relevant
in
this
yeah.
C
And
then,
if
you
go
to
the
next
slide,
so
there's
another
sort
of
more
themes
but
value
capture.
So
you
know
how
can
actually
value
of
data
be
monetized
of
code?
You
know.
Obviously,
we've
got
something
like
you
know:
the
lily
pad
project,
where,
like
you
know
some
ethical
AI
that
actually
you
know
the
original
content,
creators
or
art
creators
in
this
instance
are
being
rewarded
for
their
for
their
work.
Yeah.
A
That
one
is
it's
built
on
tablet
Willie
pad,
which
is
like
the
confusion
comes
from,
but
I
shouldn't
have
named
them
like
that.
But
I
got
the
name
water
lily
off
of
Lily
Pad,
which
is
what
the
underlying
Tech
is
for
water
lily,
but
yeah,
basically
trying
to
give
attribution
models
to
to
artists
if
anyone's
interested
in
checking
that
one
out
yeah.
C
So
so
we
are
really
interesting
that
in
terms
of
value
capture
and
then
democratizing
access
to
compute
and
training,
you
know
so
yeah.
And
how
can
you
know
this
this
be
achieved
yeah
that
them?
You
know
the
many
different
ways
to
think
about
this,
so
yeah
really
really
interesting
and
and
interesting
and
in
terms
of
how
to
solve
it
and
yeah
and
then
finally,
you've
got
the
wildcard,
obviously
and
I
think
this
hackathon
is
actually
in
person,
but
you
can
still
apply
now.
C
A
Yeah
I
I
mean
I
I
love
thinking
about
this,
because
AI
is
such
an
exploding
field
now,
but
that
also
means
there's
like
an
exploding
number
of
challenges
to
solve.
So
I
think
this
is,
you
know,
a
really
great
step
forward,
so
you
know
really
stoked
that
the
the
team
that
organized
this
has
organized
it
so
come.
A
So
I
will
hide
that
lily
pad
or
the
backyard
team
or
the
so
lilypad
is
actually
the
web
free
version
of
bakaya
we're
building
out
kind
of
these
consensus
and
Game
Theory
models
for
Lily
Pad
we're
also
going
to
be
in
Paris,
so
we'll
be
at
augment
AI.
Obviously,
we're
also
going
to
be
at
filecoin
Unleashed,
giving
a
talk
and
an
unconf
there
on
Game
Theory
and
distributed
compute
systems.
C
A
So
Lily
Pad
is
built
on
top
of
back
yeah,
so
we
all
know
that.
Well,
probably
in
case
you
don't
know,
backyard
is
a
peer-to-peer
compute
network.
You
know
working
on
the
edge
where
the
data
is
we've
seen
some
of
the
projects
built
on
it.
Super
powerful,
really
robust,
peer-to-peer,
off-chain
compute,
basically,
and
what
we're
doing
with
Lily
Pad
is
an
enabling
kind
of
this
web3
version
of
it.
So
we'll
be
able
to
run
compute
jobs
like
a
we're.
A
Building
a
compute
Network
you'll
be
able
to
run
compute
jobs
from
Smart
contracts,
or
you
know,
have
a
layer
of
payment
systems
and
you
know
be
able
to
add
your
node
so
we're
adding
the
crypto
economics
layer.
So
people
can
add
nodes,
and
you
know
you
can
be
sure
that
your
job
is
getting
run
by
their
nodes
properly.
A
So
all
the
staking
and
slashing
hence
the
game
theory
kind
of
in
distributed
systems
unconcession
and
but
at
the
moment
lilypad
v0
we're
calling
it
is
a
bridge
from
fvm
and
other
chains
to
the
backyard
compute
nodes.
So
if
you
want
to
run
a
job
say
like
stay
with
a
fusion,
you
want
to
enter
a
prompt,
and
this
is
actually
what
water
lily
runs
off.
A
Lily
Pad,
you
enter
a
prompt,
then,
basically
that's
coming
from
a
smart
contract
and
this
smart
contract
kind
of
realized
that
job
onto
the
backyard
Network,
who
then
Returns
the
job
on
chain
and
to
the
original
caller
of
that
job.
So
you'll
get
the
results
whether
they
you
know
come
back.
You
know
most
of
the
time
they
come
back
correctly
or
you'll,
get
it
canceled,
for
example.
So
that's
the
lilypad
b0,
but
we're
really
expanding
that
into
being
a
network
of
trustless
distributed
nodes.
A
A
Thank
you.
Everyone
for
joining
us
again
get
in
touch.
If
you
see
this,
you
want
to
be
on
Cod
working
group,
Live,
Jam,
rebooted
sessions,
we'd
love
to
hear
from
you.
You
can
find
us
in
the
back
of
your
slack
or
you
know,
on
our
Twitter
handles
as
well.
I
think!
That's
it
anything
to
add.
Cardiga
have
a
great
day.