►
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
C
I'm
good,
you
start
with
the.
Why
don't?
Why?
Don't
we
start
with
where
you're,
at
with
with
the
modules
II
and
I
feel
like
there
will
be
like
that.
The
flow
of
conversation
will
basically
say.
You
know
where
we
start
with
this.
What
what
we
want
to
get
across
in
this
article
chi
just
had
a
conviction,
voting
article
come
out
today
or
yesterday,
I'm
guessing
you
haven't,
seen
it
yet
see,
but
yeah
I
just
want
to
be
able
to
leverage
what's
used
in
out
and
also
yeah
I
was
just
published
on
giveth
this
morning.
C
A
So
that
makes
sense,
then,
for
me
too:
oh
god,
it
loves
it
like.
Here
we
go
okay,
so
just
to
just
to
sort
of
give
you
a
basis.
So
I
wasn't
attacking
this
from
the
perspective
of
making
the
module.
I
am
attacking
it
from
the
perspective
of
getting
a
running
model,
which
means
that,
like
a
module,
is
a
lot
of
work,
a
separate
from
getting
a
running
model,
because
what
you'll
see
is
that
a
relatively
correctly
written,
just
sort
of
like
jammed
out
CAD
CAD
model?
Has
you
know
a
little
upfront?
A
You
know,
utility
is
not
necessarily
need
to
be
in
it,
a
bunch
of
initial
condition,
construction,
which
is
just
there,
because
it
has
to
be
to
run
a
simulation
and
what
I
ended
up
doing
and
I'm
gonna
jump
I've
been
talking
to
Marcus
about
how
best
to
explain
these
things
to
the
part
where
I
actually
had
the
config
that
the
partial
state
updates.
So
actually,
logically,
the
description
of
this
model
is
I,
have
a
policy
that
generates
random
processes
to
evolve
the
network.
A
This
is
the
arrival
of
new
participants
there
at
the
suggestion
of
new
policies
and
the
arrival
of
new
funds.
Those
are
all
state
dependent,
random
processes,
meaning
they're
not
completely
random,
like
you're
not
going
to
get
new
funds
in
large
quantities.
If
the
sentiment,
slow
and
sentiment
is
one
of
the
states,
but
but
it's
still
just
primarily
a
random
policy.
A
This
is
stuff
that's
happening,
and
then
we
have
the
updates
to
the
states
as
a
function
of
those
policies
and
since
you
guys
have
sort
of
seen
the
way
CAD
CAD
breaks,
things
out,
you'll
see
a
recurrence
here.
There
are
actually
four
sub
steps,
each
of
which
is
decomposed
as
sort
of
decision
making
and
then
consequence
of
decision
making.
So
here,
because
some
proposals
have
been
passed,
we
check
progress.
A
It
means
basically
I'm
gonna
flip
a
random
variable
to
see
whether
the
project
got
completed
and
the
rate
of
completion
is
gonna
depend
on
a
couple
other
internal
variables.
But
ultimately
this
is
still
just
like
the
I
would
say
the
wrapping,
because
it's
not
the
conviction,
voting
it's
the
stuff
around
it.
But
we
talked
about
this
before
they're
sort
of
like
not
full
resolution
logic
for
the
rest
of
the
world
right,
so
a
lot
of
random
variables.
A
But
when
a
project
is
completed,
you
see
increases
in
sentiment
in
in
general
sentiment
decays
if
you're
not
completing
projects.
So
you
know
if
you
are,
if
you
tune
down
the
rate
at
which
projects
are
completed,
the
sentiment
Falls
because
there's
like
more
time
for
it
to
decay
before
it's
bumped
up
from
the
fact
that
something's
completed
and
then
we
have.
This
is
the
sort
of
one
of
the
main
blocks
for
conviction.
Voting.
It's
the
trigger
functions.
The
trigger
functions
are
basically
just
the
logic
that
we
talked
about.
A
I
have
the
function
up
at
the
top,
so
I'll
show
you,
but
basically,
if
any
particular
proposal
has
more
conviction
than
the
trigger
function
for
that
proposals,
the
gate
has
a
threshold.
Then
it
like
the
neuron
fires,
the
look
it's
past
it
releases.
There
is
a
as
I
put
a
min
time
block
on
it,
so
you
can't
release
something
in
less
than
seven
days
in
my
example,
unless
or
even
if
it
has
enough
conviction.
A
So
you
can't
just
like
put
a
proposal
up
and
pass
it
without
people
seeing
it
and
in
my
mind
this
just
means
like
you
can't
get
away
with
passing
something
before
there's
time.
For
you
know,
people
to
you
know,
look
at
it
if
they
want
to,
and
even
though
you
can't
necessarily
block
something
in
the
negative
voting
sense,
it
does
allow
for
the
normative
feedback
loops.
A
So
if
you're
doing,
if
you're
putting
something
up-
and
it's
sort
of
like
people
feel
like
it's
sort
of
against
the
spirit
of
that
particular
community
or
that
Commons
that
you're
not
gonna
like
get
it
past
them
it'll
be
like
okay,
well,
you're
doing
that,
you
know.
Maybe
it's
whatever
like
I
I'm,
not
gonna,
go
into
this
sort
of
like
politics
of
it,
but
it's
there
in
part
to
help
make
sure
that
the
normative
feedback
loops
aren't
being
bypassed
and.
C
A
A
Think
of
it
as
being
like
there's
a
there's,
a
really
critical
first
milestone
to
show
that
this
makes
sense
and
this
works,
which
is
what
I
have
now
and
that
but
I
wrote
it
in.
Like
two
I
mean
hey,
you
know,
I
read
it
in
like
two
and
a
half
days
so
like
the
fact
that
I
got
it
up
and
running
and
it
works.
How
it
was
expected
to
work
is
mostly
a
byproduct
of
the
fact
that
I
already
knew
how
the
algorithm
works,
because
I
designed
it
based
on
PhD
work.
A
So
it
was
nice
to
see
it
doing
what
it
was
supposed
to
do,
but
there's
a
lot
of
work
between
what
I
have,
which
I
think
is
good
for
narrative
and
sort
of
quick
pictures
and
little
videos,
and
something
that
would
be
right
for,
like
you
know,
forum
will
sign
off
for
something
that
would
become
the
decision
process
in
potentially
allocating
large
amounts
of
calls
that
make
sense.
So
that's
the
sort
of
description.
A
A
There's
the
last
block
is
just
updating
the
participants
decisions,
so
they
decide,
you
know
to
increase
or
decrease
their
holdings
and
cast
their
votes
and
behind
the
scenes
they
have
affinities,
which
I
can
also
make
time
varying.
The
nice
thing
about
CAD
CAD
is
like
once
you
have
a
working
thing,
it's
remarkably
easy
to
change
things
a
little
bit
and
keep
them
in
you
know,
tune
them
and
keep
them
running.
So
in
this
case,
we
had
sort
of
the
the
supply
of
tokens
rising
over
time,
which
isn't
surprising.
A
Actually,
initially
it's
going
down
because
nothing
has
been
passed.
The
system
had
to
kind
of
get
up
and
running
before
it
was
generating
positive
sentiment
because
positive
sentiment
gives
tries
to,
like
you
know,
proposals
being
so
proposals
being
passed,
gives
rise
to
positive
sentiment,
which
gives
rise
to
like
the
cailli
people.
Individual
participants
are
happy
when
their
proposals
that
they're
supporting
actually
get
passed
and
completed,
but
it
takes
a
little
time
before
that
starts
happening.
This
doesn't
start
with
anything
in
the
works.
Can.
A
Is
a
sentiment
is
an
abstracted
hidden
state
that
in
the
real
world
we
would
not
be
able
to
measure
directly
and
in
the
model
it's
just
there
to
represent
the
in
the
individual.
So
there's
two
sentiment
variables
in
this
model.
One
is
at
the
level
of
the
individual
participant
and
one
is
at
the
level
of
the
general
like
outside
population.
This.
B
A
B
A
With
a
token,
so
you
the
token
supply
is
actually
up
here
trigger
functions,
but
it
increases
because
people
are
going
and
getting
it.
So
there's
a
proxy
for
essentially
the
bonding
curve
logic
being
filled
in
here
and
right
now.
It
just
doesn't
in
it
does
not.
It
doesn't
have
an
elastic
price
or
even
really
it
doesn't
it.
This
current
version
of
the
model
that
I've
written
does
not
worry
about.
Why
or
what
it
would
cost
you
to
increase
your
supply
as
an
individual
agent.
B
A
I
mean,
but
that
goes
back
to
the
point
that
I
wasn't
yet
trying
to
make
the
I
mean,
like
I,
can
even
change
the
initial
distribution,
so
I
generated
them
randomly.
So
I
used
this
distribution
for
the
token
Holdings,
which
is
still
a
tail
distribution
for
a
reason
but
like
if
we
did
the
hatch
sale
thing
in
here,
we
would
basically
say:
okay.
Well,
let's
draw
two
sets
of
random
variables,
one
for
the
distribution
of
the
hatch,
which
will
be
much
further
up
and
one
for
the
distribution
it'll
be
like
bimodal
in
that
case.
A
But
again
this
is
all
just
in
the
setup
so
like
in
the
top.
Here
you
can
set
things
up
and
I
don't
want
to
run
super
long.
So
let
me
jump
down
to
where
I
made
plots
cuz
I.
Think
that
really
explains
what
this
does
so
I'll
go
back
up
to
the
top
of
the
plots,
so
so
just
to
give
you
a
sense
of
the
CAD
CAD
stuff,
though
so
what
happens?
Well,
CAD
CAD
throws
out
this.
A
You
know
looking
to
25,
it's
just
running
so
Oh
cuz,
I
added
all
the
states.
So
now
this
thing
is
much
got
a
lot
more
stuff
in
it,
because
I
I've
been
adding
things,
so
you
don't
need
to
worry
about
it.
The
nons
are
there
for
a
reason:
I
put
them
there
actually
for
plotting
reasons
anyway.
So
this
is
just
the
like
pic.
This
is
a
snapshot
of.
What's
in
the
data
frame,
that's
just
it's.
Just
like
a
table
structure.
We've
got
plots
of
supply
and
sentiment
and
funds.
A
A
The
network
X
data
set
so
I'm,
like
the
network,
X
state
variables,
so
I'm,
basically
plucking
the
data
out
of
the
graph
object
using
these
lambdas
and
putting
them
in
more
traditional
like
well
like
records
and
in
the
they're,
like
they
become
sort
of
like
nest
tables
with
nested
values,
because
I'm
putting
I'm
putting
numpy
arrays
in
in
elements
of
tables,
and
so
all
of
this
is
just
collecting
up
the
data
from
the
simulations
I,
because
there
are
four
sub
steps
per
time.
Step.
A
But
I
can
actually
also
layer
in
proposal
by
proposal,
so
that
certain
proposals
are
conflicting
and
if
one
passes
that
the
other
ones
who
are
for
the
same
thing,
sort
of
lose
their
votes,
because
no
one
wants
both
and
I
can
also
put
progressive
participants
by
participant
edges
that
represent
sort
of
mutual
influence
also
relatively
easily,
because
of
the
construct
and
I
would
expect
both
of
those
things
to
be
in
there
for
the
for
their
actual
engineering
work.
Even
though
they're
not
in
here
for
the
initial
sort
of
demonstration.
Does
that
make
sense?
You.
A
Social
network
effects
in
them
in
the
participant
by
participant
and
sort
of
counter,
like
sort
of
competition,
graphs
and
conflict
graph
between
the
proposals,
so
that
a
proposal
that
is
passed
could
render
a
previous
extremely
unlikely
to
pass,
even
though
it
doesn't
technically
remove
it
from
the
list.
It
would
be
implying
that
you
know
it
doesn't
make
sense
to
have
both
so.
C
A
C
A
C
A
Technically,
you
could
still
pass
both
because,
if
there's
enough
conviction
to
pass
the
second
one,
even
after
the
first
one
passes,
it
just
means
that
the
community
is
chosen
choosing
to
de-risk
something
by
by
funding
two
separate
initiatives
on
the
same
topic,
which
is
nothing
wrong
with
that,
because,
technically
we're
not
guaranteed.
This
is
just
funding.
It's
not
the
outcome
so,
like
you
could
fund
two
projects
to
do
almost
exactly
the
same
thing
and
that's
a
like
a
parallelization.
C
A
A
So
I
mean
I
again.
This
is
mostly
just
I'm
talking
to
you
guys
about
the
perspective
from
what
you
what
I
have
here
versus
where
we're
going,
and
so
things
like,
we
will
need
proposal
to
proposal
underlying
graphs
and
participant
to
participant
lying
grafts
in
order
to
represent
the
problem.
But
it
also
to
be
clear
that
a
big
chunk
of
this
is
actually
computational
social
science.
The
engineering
part
is
just
using
the
computational
social
science
level
or
layer
and
just
changing
the
outfit
so
that
you
can
say
well
what,
if
that's
like
this?
A
A
Because
participants
just
I,
don't
know
I
guess
this
is
just
not
common
knowledge.
So
if
you
do
a
study
of
group
decision
making
the
actual
social
influence
everything
up
people's
influence
on
each
other
basically
makes
their
private
signals
harder
to
extract.
So
if
you
believe
that
the
private
signals
of
the
participants
in
a
system
like
this
have
the
information
that
you
need,
but
then
you
let
them
basically
gossip
with
each
other.
These.
A
Process
basically
completely
screws
up
the
underlying
signal
extraction.
So
from
a
signal
processing
level,
it
actually
makes
well
I
mean
I,
can't
necessarily
say
it
screws
it
up.
It
makes
the
answer
very
different
and
so
understanding
that
that
is
that
the
signal
processing
level
having
a
very
large
effect
it
wouldn't
make
sense
to
ignore
it.
You
need
to
be
able
to
like,
oh
that
bottle
yeah,
because
this
is
a
graph
data
structure.
It's
not
that
hard,
but
this
is
a
lot
of
stuff.
That's
not
going
to
be
represented
in
in
any
smart
contracts.
A
It's
a
it's
a
lot
of
the
difficulties
in
the
representation
of
the
world
that
you
use
in
order
to
validate
in
test
the
smart
contracts
which,
again,
technically
speaking,
the
smart
contract,
is
going
to
be
able
to
do
exactly
what
it's
supposed
to
be
able
to
do.
The
real
question
is
whether
it
serves
the
needs
of
the
people
and
you
need
to
have
notions
of
what
that
means,
and
here
that's
gonna
mean
things
like
you
know.
The
decisions
are
in.
A
My
wine,
you
know
with
the
affinity,
is,
for
example,
that,
like
peoples
actually
actual
preferences,
you
know
private
signal
preferences
are
being
expressed
in
a
meaningful
way.
We're
never
gonna
be
able
to
say
this
was
the
right
answer,
and
this
was
the
wrong
one,
because
there's
many
different
ways
of
aggregating
up
those
low-level
preferences.
But
if
you
combine
the
sort
of
method
of
building
out
a
set
of
underlying
preferences
and
then
the
set
of
outcomes,
you
can
measure
various
sort
of
social.
A
A
And
I'm
gonna
try
to
go
a
little
quicker
because
I
really
want
to
give
you
the
floor
before
we
run
out
of
time.
This
is
the
this
actually
I'll
make
a
really
important
point.
The
thing
that's
really
hard
and
is
awesome
about
CAD
CAD
is
the
whole.
The
evolution
in
time
is
where
all
the
interesting
stuff
is
happening.
It's
not
necessarily
about
any
one
decision.
It's
about
how
the
system
is
sustainable,
right,
its
regenerative
regenerative.
So
here
we
see
the
seven
candidate
proposals
at
the
beginning.
A
You
pass
some
of
them
would
they
become
active
and
when
the
active
proposals
are
completed,
they
become
completed
and
you
see
accumulating
completed
projects
over
time,
but
you
actually
see
a
continued
like.
There
are
candidates
there
pretty
much
continuously
because
new
ones
are
arriving.
So
what
we're
really
trying
to
do
is
convert
candidates
into
active
and
active
into
completed
and
accumulate
output
right.
We
want
material
output
from
this
community,
not
just
funding.
A
That's
the
the
catch
with
a
lot
of
the
funding
you
know,
Dow's,
is
that
they're
sub
optimized
in
the
sense
that
they're
designed
to
fund
things
they're
not
actually
designed
to
fund
things
that
produce
things
to
fund
things
in
the
future.
Like
you,
don't
close,
the
loop
and
I
have
the
other
diagram
that
I
showed
you
before,
which
I'm
actually
tempted
to
snap
into
here
with
the
green,
the
yellow
and
the
blue,
showing
the
like
by
loop
centered
on
production.
On
the
actual
you
know,
completion.
B
A
B
A
To
give
it
okay,
so
here
I
renormalized
by
the
funds
so
instead
of
count
and
what
you'll
see
is
that
because
the
system's
accumulating
funds,
actually
the
size
of
the
projects
is
growing.
So
the
candidate
went
up
a
lot
because
later
there's
a
lot
of
bigger
proposals
being
made
they're
just
not
passed
right
because
they
have
to
be
outstanding,
otherwise
they
wouldn't
be
in
candidate
funds.
The
people
are
asking
for
more
money.
A
A
Yeah
yeah
I,
so
I
set
this
up
in
base
okay,
so
it's
at
least
logically
in
days
well,
yeah
and
obviously
a
lot
of
the
processes
can
be
retuned
with
timescales
in
mind.
Just
remember,
this
whole
thing
is
pipes
so,
like
a
lot
of
the
specifics
of
what
you
see
are
just
a
realization
from
a
run
that
I
ran
and
I
got,
we
have
a
lot
of
control
over
the
initial
command,
the
the
logics
through
which
it
evolves
and
the
logics
through
which
people
make
decisions.
A
We
just
wanted
to
show
as
quickly
as
possible
that
the
properties
that
we
said
the
algorithm
has
in
time
like
are
actually
the
properties
that
it
has
that,
like
as
a
demonstration,
and
so
this
is
just
conviction,
it's
not
too
informative.
It's
kind
of
squiggly
I
made
the
chart
that
you
had
asked
for.
So
this
is
the
share
of
funds
requested
by
proposal.
You
can
see
it
going
down
largely
because
the
funds
are
going
up,
but
the
specific
plot
that
you
know
you
and
I
had
talked
about
was
this
one.
A
A
To
not
the
other
challenge
here
is
that
there,
the
the
color
cycle
so
I
had
some
notes
in
this
to
try
to
figure
out
how
to
get
a
different
color
palette,
because
it
like
that
blue.
It
cycles
through
1
2,
3,
4,
5,
6
7,
like
maybe
10
colors
and
then
cycles
again
so
like
this
even
weirdly
enough.
This
orange
and
that
orange
are
different
and
it's
so
it's
really
hard
to
speak
because
they're
overlapping.
The
reason
why
they're
they
do.
A
B
A
C
B
A
I,
so
in
this
one
I
rescaled
it
so
on
this
one.
It
was
time
in
this
one,
it's
age,
so
the
x-axis
switched
from
natural
time
to
renormalize
so
that
each
puzzles
time
starts
when
it
was
born.
And
now
you
can
see
this
cluster,
that's
what
you're
talking
about
they're,
the
ones
that
passed
on
their
seventh
day.
Okay,.
A
But
you
can
also
see
that
this
green
one
here
was.
It
took
the
entire
it
took
its.
It
took
the
entire
simulation
two
paths.
Basically,
if
it's
it's,
it
started
in
the
beginning,
because
there's
only
200
days
in
this
simulation,
so
it
had
to
in
order
for
it
to
line
up
from
there
there.
It
means
that
it
literally
was
one
of
the
day
zero
proposals
and
it
and
it
doesn't
pass
till
the
end
and.
B
A
C
A
Yeah,
so
all
that,
so
what
I
was
saying
is
that
behind
the
scenes,
this
is
a
essentially
a
graph
database
in
state.
So
when
I
this
thing
that
I
did
all
this
mess
here
was
to
generate
this
plot,
which
is
the
thing
I
made.
The
video
from
and
I
used
that
data
to
make
this.
These
are
sized
by
the
funds.
A
These
are
by
the
funds
requested,
and
these
are
sized
by
Holdings.
I
might
want
to
renormalize
that
so,
like
I
could
scale
up
the
the
size
of
the
the
these
were
when
I
first
did
it,
they
were
really
overlapping.
You
know,
I
could
talk
to
data
vis,
people
and
sort
of
clean
these
up
and
even
make
like
a
histogram
that
points
to
the
left
or
something
if
you
got
like
a
really
high
end
date
of
is
guy
but
like.
A
Basically,
this
is
the
size
of
the
holdings
of
the
participant,
and
so
this
is
the
hundred
hundred
and
twenty
odd
participants.
The
lines
are
they're
they're,
they're
voting
weights
and
the
reason
why
there
was
a
video
is
because
essentially
it's
just
these
images
down
here
over
time.
So
this
was
time
slot
0.
Is
this
little
tiny
proposal
it's
being
supported
by
a
lot
of
people,
but
it
takes
till
the
seventh
day
to
pass
and
there's
not
that
much
funds
at
the
beginning.
A
So
you
know
you
just
see
that
charge
up
and
then
it
turns
yellow
because
it
passed,
but
you'll
also
notice
that
the
lines
start
changing
too.
So
those
are
what
people
are
voting
for
and
I
mean
it's
easier
to
see
in
the
video
the
this
is
200
screenshots,
basically,
that
I
use
to
make
the
video,
but
later
on,
you
can
start
to
see
things
like
this
happening
so
like
this
graph
is
changing
in
time,
I'm
showing
you
the
edges
only
for
the
votes
and
only
with
the
weight
of
the
votes.
So
the
really
light.
A
A
A
But
I
mean-
and
you
know
this
thing
just
keeps
going
that's
what
this
is.
This
is
what's
in
the
video
and
I
can
make
them
again
relatively
easily.
The
hardest
thing
is,
it
would
be
really
nice
if
I
had
somebody
with
a
little
more
experience
who
could
directly
make
the
video
cuz
I
hacked
it
I,
just
downloaded
all
the
pictures
and
put
them
in
an
iMovie
I
know
that.
A
A
B
A
I
would
love
to
chat
with
him
at
some
point,
because
I
can
make
I
mean
I
can
make
things
like
this
pretty
quickly,
but
it's
less
about
the
mean
it's
more
I
was
thinking
more
in
terms
of
the
programming,
because
I
was
looking
it
up
and
you
can
do
a
complete.
You
don't
need
to
do
like
iMovie.
You
can
literally
write
the
right
name
block
of
code
and
then,
instead
of
dropping
means,
it
will
compile
them
into
a
video.
I
did
a
really
happy
thing
where
I
printed
them
out
saved
them
using.
A
You
know
a
like
a
line
of
code
that
saves
them
to
disk
and
then
I
took
the
files
and
loaded
them
into
a
video,
but
you
can
do
all
that
in
line
it
was
with
this
like
animation
stuff
that
I
don't
really
know
how
to
use
anyway.
So
that's
the
model
that
I
created
so
far,
I'm
gonna
show
you
quickly.
One
other
thing
before
I
wow
God,
stop
stop
it.
How.
A
So
I
wanted
to
show
this
other
screen
quickly.
So
this
is
this.
This
is
probably
gonna,
be
an
Abbie's
article,
but
this
is
a
rough
hack
of
a
drawing,
not
a
final
drawing
I,
don't
know
if
she's
gonna
use
it,
but
this
is
really
important.
In
fact,
every
aspect
of
what
we're
doing
boils
down
to
recognizing
that
that
actually
the
thing
driving
these
systems
is
the
real
value
produced
and
the
thing
the
real
value
produced,
the
real
work,
the
milestones,
the
stuff
being
done
by
the
people
which
isn't
really
the
Dow
itself.
A
It's
the
people
using
the
comments
they
get
funds
from
one
side
and
then
they
get
decision-making
from
the
other.
So,
like
funds
flow
in
through
the
funding
pool,
and
while
the
proposals
are
the
are
being
you
know,
sorting
those
funds
intuitively,
you
can
think
of
this,
like
a
labor
and
a
capital
being
on
the
same
page,
because
the
decision-making
loop
is
the
thing,
the
decentralized
government
side,
governance,
side,
the
the
green
drawing
here.
This
is
where
you
can
think
of
it.
A
As
the
token
it's
the
it's,
the
community
deciding
what
to
do
and
the
funding
that's
required
to
do
it.
So
you
so
you
have
the
sort
of
capital
coming
in
from
the
right
and
the
labor
from
the
left,
saying:
here's
what
we're
gonna
do
and
then
the
yellow
bar
is
actually
doing
it
and
it
produces
outputs
and
in
our
system
that
output
resolves
to
valuable
outputs
that
result
in
future
funding
and
tokens
which
represent
future
influence.
A
A
It's
actually
primarily
centered
on
the
actual
work
being
done,
and
it's
only
the
fact
that
that's
happening
that
leads
to
people
caring
about
having
tokens
for
future
decision-making
and
people
actually
caring
enough
to
put
funds
in
so
and
that's
inherently
outside
of
the
of
the
automation
of
a
smart
contract.
It
has
to
do
with
what
people
do
given
what
they
did.
A
So
that's
this
right
side
of
this
drawing
and
I
I
think
that
it's
really
important
to
remember
that
all
of
this
is
just
a
sort
of
facilitate,
or,
for
this
sort
of
you
know,
production
line
this,
like
horizontal
line
from
sort
of
milestones,
complete
to
material
outputs.
That's
the
actual
purpose
of
this
system
is
to
help
these
people
coordinate.
A
To
do
this
thing
using
these
funds,
so
the
yellow
bar
is
actually
the
center,
and-
and
this
is
one
of
the
reasons
why
I
criticize
so
strongly
a
lot
of
the
things
that
have
come
that
are
being
done
and
discussed
is
not
that
I
think
that
they're
bad
parts,
but
the
whole
system
ends
up
like
it's
next.
This
is
the
actual
production.
A
If
that
fails
or
is
done
poorly
or
is
managed
poorly
or
anything,
it
means
that
there
won't
be
a
continuous
source
of
funds
because
it
comes
from
people
appreciating
or
paying
for
or
granting
the
continued
labor
and
likewise
there'll
be
no
demand
for
the
tokens,
because
the
decision-making
over
that
is
only
valuable
if
the
thing
itself
is
valuable,
and
so
you
actually
put
yellow
in
the
middle
and
obviously
this
drawing
is,
is
not
the
same
as
the
little
one,
because
this
is
more
of
us
more
actually
of
a
value
flowspec.
So
it
depends
on.
A
You
know
how
things
actually
fit
together
it.
So
the
drawing
is
clear.
The
actual
stylized
sort
of
what's
happening
is
this
and
conviction
voting
is
essentially
the
in.
If
you
think
of
the
bonding
curve
is
just
the
boundary,
then
you
can
almost
think
of
the
bonding
curve
over
here.
I,
don't
really
know
how
to
one-to-one
this
to
the
bonding
curve,
but
I
would
say
this
is
the
conviction,
voting
stuff.
A
This
is
giveth,
and
if
we're
really
talking
about
it
than
three
components,
I
would
say
you
know,
maybe
this
maps
to
the
Augmented
bonding
curve,
because
the
Augmented
bonding
curve
has
the
funding
pool,
but
all
in
all
the
this
system
is
centered
on
the
leverage
is
actually
on
the
on
the
production
itself,
not
on
the
token
and
not
on
the
capital,
like
you
need
those
two
things,
but
it's
like
the
it's
like.
Those
are
inputs
and
outputs
of
this
central,
like
pillar
tool.
C
A
Notes
and
I
mean
to
be
clear,
like
I'm,
also
like
I've
got
regular
meetings
with
people
from
Sherman's
group
and
I've
talked
to
her.
Some
that's
one
of
the
reasons
I
recommended
her
I
read
her
I,
like
was
one
of
the
reviewers
for
the
chapter
on
behavioral
economics
and
mechanism
design
in
her
book
I'm
like
I'm,
feeling
really
good
about
the
knowledge
and
expertise.
A
That's
coming
from
the
more
economic
side,
which
is
where
you
see
the
discussion
of
goods,
for
example
like
what
does
it
mean
for
to
have
a
common
zone
good
in
the
blockchain
era,
and
that's
kind
of
where,
as
pushing
a
beyond
the
cyber
physical
systems,
because
I
still
think
there's
a
lot
of
ignorance
about
the
parts
of
the
system
like
what
did
you
produce?
Who
owns
it?
And
what
does
that
mean
to
the
people
who
would
like
participate?
And
if
you
ignore
that
stuff,
you
really
aren't
closing
the
loop
all
right.
C
C
C
A
So
so,
if
you
want
to,
if
you
want
to
be
the,
if
you
want
to
be
the
person
who's
taking
ownership
of
the
proposal,
because
there's
gonna
be
this
downstream
notion
right
that
the
proposal
gets
funded
and
you're
gonna
have
an
initiative
and
I
mean
there's
a
place
for
Grif
to
step
in
and
talk
about
how
giveth
can
or
should
or
does
work,
but
like,
ultimately
that
if
the
trigger
valve
releases
funds
to
an
initiative,
some
address
in
some
person,
some
identity
likely
has
stewardship
over
that
proposals.
Completion.
A
And
so,
if
you
are
the
proposer
of
a
particular
proposal,
I
think
you're
in
a
different
category
than
a
supporter
and
I
was
saying
stake,
because
I
was
imagining
that
that
Staker
was
actually
locked
in
and
if
they
wanted
out,
they
would
either
need
to
get
someone
else
to
take
like
take
ownership
of
that
proposal,
or
they
would
have
to
take
it
down.
That
you
can't
have
unknown.
You
can't
have
ownerless
proposals.
They
have
long
to
some
community
member
who
asserts
that
by
staking
yes,.
B
So
there's
like
the
proposing
it
to
the
DOE
and
proposing
it
to
give
us
which
are
two
different.
So
there's
like
the
stake
of
like
getting
it
to
show
up
in
the
voting
interface,
but
then
there's
allocating
your
tokens
and
voting
basically
yeah
and
that's
the
I.
Guess
you
call
it
your
call
on
it.
Assert
tokens
cert.
A
Yeah
I
mean
Express
preferences
like
we
can
play
around
with
the
words
there's
a
little
bit
of
like
a
design
freedom
there.
We
know
what
it
means.
Technically,
it
means
that
you
have
an
onion,
have
an
unrestricted
right
to
taesik
aliy,
take
your
balance
and
allocate
it
against
your
preferences
and
that
it
will
accumulate
conviction
for
the
proposal
according
to
the
conviction,
voting,
and
you
can
move
it
whenever
you
want,
but
your
the
conviction
will
will
vary
based
on
those,
so
you
just
spend
it
and
the
system
is
going
to
as
part
of
that
action.
A
When
it,
you
know
it's
going
to
be
able
to
update
itself
to
to
account
for
the
different
accumulation
right.
So,
like
it's
gonna
discharge
over
time,
it's
not
even
gonna,
take
away
conviction,
it's
just
gonna,
accumulate,
dis,
lower
and
thus
potentially
go
down
right.
The
capacitor
discharging
thing
I
will
say
that
there's
gonna
be
a
step.
A
That's
gonna
be
really
important
and
it's
on
my
to-do
list
to
write
up
more
about
temporal
states
and
smart
contracts,
because
what
I've
been
you
know,
I,
don't
wanna,
say
fighting
with,
but
like
trying
to
make
sure
that
people
understand
is
like
a
consistent
way
to
deal
with
time.
Varying
things
and
I
have
some
thoughts
on
sort
of
predicate
state,
update,
logics
that
basically
just
say.
Whenever
someone's
to
transact
on
a
contract,
then
whoever's
the
next
actor
bears.
A
You
know
some
cost
of
what
I
will
call
the
predicate
state
update,
which
means
certain
amount
of
blocks
have
passed.
Therefore,
a
state
update
happens
prior
to
your
action,
and
you
can
always
check
the
predicate
state
update
with
an
off
chain
call
by
running
that
predicate
state
update
function
but
predicate
state
update
just
means
the
the
real
state
of
the
contract
is
the
prior
state
and
whenever
you
go
to
act
on
it,
anything
that
quote
happens
between
the
last
state
and
now,
as
a
result
of
just
time,
is
resolved
by
a
predicate
state
update.
A
This
is
a
I'm
I've
been
talking
about
it
as
a
reference
architecture
thing
to,
though,
at
this
point,
because
I've
run
into
it
a
lot
and
I
have
a
pretty
good
sense
of
how
to
specify
it.
As
a
generic
methodology
says,
whenever
you
define
things
as
systems
like
in
a
systems
engineering
way,
if
they're
they
can
be
purely
state
dependent
or
they
can
be
state
state
and
time
dependent
and
in
the
blockchain
systems.
A
C
A
C
This
was
perfect.
This
is
perfect
and
just
go
through
some
of
the
main
points.
I
was
looking
to
cover
and
maybe
clarify
a
few
of
those
or
drop
a
few
or
add
a
few
whatever,
and
so
we're
looking
to
create
pretty
much
a
component
explainer
series,
I'm
kind
of
like
high-level,
taking
the
math
into
English,
explaining
the
features
and
functionality
rather
than
the
the
math
and
the
logic
behind
it.
So
yeah
this
will
be
the
contingent
voting
as
the
continuous
is
making
alternative
a
couple
of
alternative
titles.
A
C
A
C
A
To
end
up
having
a
whole
nother
discussion
about
TC
RS
and
like
ontologically,
correct
framing
of
coking
curated
registries,
because,
frankly,
I
have
a
lot
of
opinions
on
that
and
they're,
not
as
directly
aligned
with
I
mean
conviction.
Voting
is
more,
is
a
different
kind
of
decision-making.
Well,
we'll
just
leave
a
pause
and
come
back
to
TC
RS,
because
I
think
we
can
talk
about
them.
Just
I
don't
want
to
mix
this
in
with
it
sounds.
C
Good
I'm,
okay,
so
high-level
explanation
of
conviction
voting.
Ultimately,
we
are
trying
to
order
proposals.
That's
not
really
ordering
I
know
you
mentioned
that
it's
not
a
strict,
a
versus
B
versus
C,
but
it's
trying
to
use
continuous
preference
to
order
those
proposals
according
to
preference
broadcast
remembers.
How
do
we
decide
as
a
collective
how
to
allocate
funds?
Basically
the
modules
it's
between
the
bonding
curve
and
the
giveth
DAP
we're
approved
proposals
are
funded.
We
want
humans
to
remain
in
charge
of
the
socio-technical
systems
were
bringing
into
existence.
C
Humans
are
social
sensors
conviction
is
fusion
or
aggregation.
I'll
reference.
Your
paper
here
of
those
inputs
towards
necessary
proposals
and
system
updates.
Quick
note
here
that
it's
not
just
strictly
ordering
proposals.
Some
of
this
will
definitely
be
tweaked.
I
also
had
a
note
from
you
saying
that
this
was
kind
of
like
machine
learning
with
humans
unsupervised,
one
versus
all
classifier
were
the
inputs
or
preferences
expressed
by
community.
A
Ontologically
right
Kim,
you,
man,
preferences,
are
like
continuous
time
varying
messy
things,
but
we
need
discrete
concrete
outputs,
and
so
our
neuron
firing
is
a
really
good
example
of
a
system
that
has
this
sort
of
messy,
noisy,
continuous
time
varying
thing
that
gets
converted
into
a
discrete
event,
like
a
discrete
binary
event
like
this
happening,
and
so
we're
modeling
the
off
off
something
that
has
a
simple
logical
similarity
in
that
it's
sort
of
noisy
and
messy
and
varying.
But
we
need
a
discrete
event.
Output,
cool.
C
C
Now
do
I
do
remember.
I
just
wanted
to
confirm
that
was
the
the
correct
terminology,
I'm
cool,
so
I
was
going
to
put
in
a
quick
paragraph
on
how
this
differs
from
what
we
do
today
now.
Maybe
this
is
more
of
a
TCR
chat.
I
was
just
looking
to
go
into
kind
of
some
of
our
I
mean
we
don't
actually
have
to
save
problems
with
TC
hours.
It
could
be
problems
with
time
time
based
bonuses.
A
I
was
general
here
to
compare
I
would
I
would
I
would
I
would
address
this
similar
the
differences
from
time
based
time,
boxed
voting
and
the
challenges
with
things
like
the
Eragon
votes
as
the
pain
point,
and
then
we
could
make
us
like
two
or
three
sentences
like
hey,
like
there's,
also
some
relations
to
TC
arse,
but
we're
gonna
explore
you
know.
Tc,
RS
and
biomimetic.
Tc
are
alternatives
in
another
article
and
so
like
I.
A
Think,
because
we're
gonna
have
to
just
go
into
a
lot
more
than
we
can
handle
here
to
properly
address
the
basically
I'm
going
to
end
up
giving
you
a
breakdown
of
what
TCR
czar,
what
they
can
and
should
be,
and
unfortunately
that's
just
a
bunch
more
work.
Yeah
like
there
I
we've
talked
about
it
a
few
times
and
that
does
yeah.
We.
C
A
Buying
is
something
we
want
to
address
last
minute,
both
swings.
We
would
want
to
address
voter
apathy.
We
would
want
to
address
awful
UX
I,
don't
know
about
this
last
ones.
Time
interval
too
long
arbitrary,
set.
Look
at
that
yeah,
so
the
the
last
couple
seem
like
they're
consequences
of
the
first
few,
so
I
don't
know,
I
mean.
A
A
A
A
A
The
other
thing
that
I
was
gonna
say
is
that
your
preferences
being
broadcast
is
something
that
you
can
and
I
guess.
It
depends
on
the
UX
layer
stuff
but
like
because
you
don't
have
to
vote
in
a
at
a
fixed
time
box.
Essentially,
you
can
periodically
just
go
check-in,
set
your
preferences
and
walk
away,
and
if
you
do
that,
once
a
month,
you're
still
actively
participating,
you
are
not
being
forced
to
pink
attention
constantly
to
find
out
whether
or
not
there's
something
you
should
be
voting
on
and
missing
it.
A
So,
in
a
sense,
the
the
cognitive
cost
of
participation
is
lower,
because
if
you
check
in
on
your
own
schedule
that
you're
never
gonna
get
a
situation
where
it's
like.
Oh
hey,
there's
a
vote
in
two
weeks
and
then
you're
like
putting
it
on
your
calendar
or
whatever
you're.
Actually
just
going
oh
cool
I
like
that
one,
maybe
I'll
my
settings
around
and
then
I
don't
have
to
pay
attention
to
it
again.
Yeah.
B
And
I
also
think
that
we
need
to
talk
about
implementation,
detail
plumbness,
just
a
little
bit
talking
about
implementing
ERC
made
a
date
and
being
able
to
have
constant
like
if
you
get
new
tokens
there
are.
Your
preferences
are
being
allocated
for
those
new
tokens
already
right
and
and
it's
sort
of
like
Arab.
A
Holdings
that
that's
part
of
the
reason
that
the
assert
can
make
sense
right
if
you're
soakin,
your
your
time
varying
Holdings
are
totally
fine,
because
your
preferences
are
expressed
as
a
share
of
your
holdings.
Yeah,
yeah
and
I
mean
behind
the
scenes.
We
have
to
have
to
figure
out
how
to
make
sure
that
they're
rounding
errors
and
stuff,
but
like
practically
speaking,
it'll,
be
fine.
A
Also
pretty
straightforward
right,
because
if
your
list,
if
your
allocation
is
a
percent,
is
oppressor,
is
a
weight,
if
you
think
of
it
as
a
weight,
I
have
power
one
two,
three,
four:
five
over
the
list
of
things
I
support
and
then
your
actual
percentages
resolved
to
sort
of
what
percent
of
your
total
weight
you
allocate
for
your
tokens.
Then
all
you're
doing
is
popping
the
thing
that
passed
out
of
that
list.
Yeah.
B
A
B
A
You're
gonna
have
some
state
variables
that
aren't
immediately
like
they're,
like
once
removed
that
are
gonna
get
updated
like
I,
suppose
I
send
you
some
funds
like
potentially
I'm
gonna
need
to
do
I
mean
we
ask
we
get
to
figure
out
when
and
when
it
should
and
should
actually
trigger
the
predicate
State
update
but
like
there's
definitely
gonna,
be
some
interesting
things
in
terms
of
how
we
track
those
sort
of
state,
evolutions
and
compute.
These
things,
I
really.
B
A
B
A
B
C
A
Think
we're
broadly
because
of
the
nature
of
this
contract
in
this
system,
we're
going
to
be
broadly
relying
on
that's
one
of
the
reasons
why
I
think
we
may
actually
want
to
take
a
time
out
to
write
something
about
predicate
state
logic,
because
that
what
I'm,
what
I'm
describing
is
the
general
reusable
concept
formally
for
what
you're
talking
about
it's?
Something
where
you
could
prove
that
the
system
worked
as
intended,
essentially
mathematically
by
showing
them.
There's
no
thing
that
you
can
do
that.
Doesn't
force
computation
of
the
predicate
state
that
would
violate
that
state.
What.
B
I
would
say
is
that
I
know
that
we
I
give
it.
We
are
implement
something
similar
in
the
death
and
I.
Would
during
this
talk,
I
was
cruising
the
ip's
looking
for
something
along
these
lines.
I,
don't
know
that
I
see
one
but
I'm
sure
that
in
the
counterfactual,
like
crew
I
mean
I
got
when
I
asked
Jordi
and
I
want
to
ask
some
counterfactual
guy
if
they
know
of
something,
because
I
don't
want
to
write
this
article,
when
there's
gonna
be
some
other
article
that
we
should
have
brought
yeah.
A
Yeah
and
I'm
not
talking
about
writing
the
article
from
the
perspective
of
the
implementation.
I
would
be
just
as
happy
to
cite
the
person
who's
doing.
Who
has
who's
done
things
that
follow
this
pattern,
but
I'm
talking
about
reference
architectures,
not
so
much
about
like
and
there's
a
level
of
abstraction
beyond
anything
that
I've
seen.
A
There
are
cases
of
counterfactual
type
things,
but
there
is
a
very
sort
of
rigorous
and
formal
statement
that
says
you
can
always
do
the
following
thing:
if
you
can
ever
sort
of
factor
your
logic
in
this
way,
you
can
always
do
this
and
it
handles
a
class
of
problems.
So
what
I'm
basically
saying
is
we're
gonna
run
into
this
class
of
problems
a
lot?
We
know
how
to
solve
this
class
of
problems
by
factoring
it
like
this
and
hey
by
the
way.
We
know
this
works,
because
these
other
guys
did
a
version
of
this
here.
A
C
No,
no
good
good,
I'm
cool,
so
yeah
I
guess
for
some
of
these
do
definitely
fit
into
you
know.
Whale
dominance
is
kind
of
boat
buying
or
you
know,
influence
of
cash
on
or
wealth
on
system.
I
just
wanted
to
address
some
of
the
other.
You
know
like
these
arbitrary,
lock
periods,
the
multiple
transactions
I.
Guess,
that's
not
so
much
error.
C
C
A
That
workers,
you
let
me
know
specifically,
though,
about
plotting,
because
I
will
generate
images,
is
that
that
code
is
on
the
conviction,
repo
like
if
you
flip
through
it,
and
get
a
sense
of
what
you
want.
Then
let
me
know-
oh
my
god
sever.
Oh
it's,
this
cleaners
they're
here
every
week,
basically.