►
From YouTube: TEC Co-Lab - Praise System Data Analysis 30-05-2021
Description
🙏 Thank you for watching! Hit 👍 and subscribe 🚩 to support this work
🌱Join the Community🌱
on Discord https://discord.gg/uM4ZWDjNfK
or say hello on Telegram https://t.me/tecommons
Join the conversation https://forum.tecommons.org/
Follow us on Twitter: https://twitter.com/tecmns
Learn more http://tecommons.org/
A
Intentions,
I
have
two
intentions,
or
maybe
three
so
first
of
all,
my
friend
saul
is
here
who
does
3d
modeling
and
we've
identified
a
really
cool
potential
way
to
look
at
the
the
word
count
of
the
most
common
words
that
came
out
of
the
praise
data
to
render
that,
as
like
a
three
word
sphere,
so
excited
for
that
it's
a
fun
project
more
on,
like
the
front
end
and
design
side
of
things.
Number
two
is
just
to
announce
to
the
community
that
my
outputs
from
last
time
were.
A
One
kind
of
the
word
count
thing
that
some
people
might
want
to
check
out
and
it
could
be
expanded.
Number
two.
I
started
a
framework
for
using
doctorvec
to
spatially
embed
the
praise
column
and
that
can
be
picked
up
by
anyone
here.
Who
might
want
to
take
a
look
at
that
and
do
that
work?
I
don't
I'm
not.
I
don't
plan
on
getting
back
to
it
right
away
because
of
the
third
thing
that
I
wanted
to
mention,
which
is
what
I
think
I
personally.
A
If
what
I
want
to
tackle
today
to
be
most
effective
is
I
would
really
like
to
start
making
ground
on
the
idea
of
looking
at
the
discount
rows,
so
they
get
applied
and
try,
because
there's
a
bit
of
data
cleaning
that
would
have
to
be
done
complementary
to
what
octopus
has
done
in
terms
of
just
isolating
those
discount
rows
and
then
being
able
to
line
them
up
so
that
I
could
reverse
that
and
quantify
how
much
the
discount
has
been
to
those
who
are
on
sort
of
like
a
salary
style
with
the
common
stack
or
the
tec,
and
do
some
quantification
on
the
balance
there,
and
I
I
sort
of
have
some
intuition
around
a
few
equations,
where
maybe
we
can
equate
potential
outcomes
of
the
dollar,
the
dollar
amount
of
a
tech
token,
and
then
how
that
looks
compared
to
the
salaries
that
were
paid.
A
So
I
think
that's
a
really
interesting
one.
I'd
like
to
make
some
ground
work
on
that
myself,
but
those
are
the
three
points
having
some
fun
with
a
word
sphere,
with
saul
with
some
3d
modeling
number
two
is
the
doctor.
Vec
work
can
be
picked
up
by
anyone.
Any
curious
data
scientist.
It's
I
kind
of
started
it.
I
think
it's
a
good
project
and
number
three
I'm
going
to.
I
would
love
to
lay
the
groundwork
for
equating
the
the
costs
of
governance.
B
Cool
yeah
intentions
eager
to
see
sort
of
where
the
the
the
data
analysis
stuff
is
at
and
hear
about
any
insights
that
have
been
gleaned
thus
far
yeah
curious
to
touch
base
and
see
kind
of
sort
of
like
time
check
sense
making.
In
terms
of
like
what
kind
of
80
20
actionable
insights
we
can
take
that
are
also
mindful
of
timelines
that
the
tec
is
working
towards
and
I'm
curious
how,
because
I
I
feel,
like
we
have
a
couple
of
work
streams.
B
One
is
like
for
future
evolutions
and
modifications
and
upgrades
of
the
praise
system
moving
forward
and
they
and
another
may
be.
You
know
more
sensitive
towards
pre-hatch
considerations,
so
just
curious
to
sort
of
pick
apart.
You
know
what
are
the
80
20
things
for
anything,
that's
time
sensitive
and
then
also
you
know,
looking
towards
the
phrase
system
as
a
gradable,
you
know
process
moving
forward
and
how
we
can
make
sure
that's
suiting
the
needs
of
the
people
that
are
in
it
distractions.
B
None
on
my
side
and
I
will
pass
it
on
to.
I
was
a
little
bit
late,
so
I'm
not
sure
who's
gone
yet,
but
I'll
pass.
It
on
to
johan.
C
C
A
C
C
C
Second,
hey:
do
you
mind
just
coming
back
to
me
in
a
sec,
I'm
just
gonna
pull
up
the
example
of
the
idea.
Whoever's
next
can
go
just
quickly
and
then
circle
back
to
me.
E
E
Thanks
septimus
yeah:
my
intention
is
to
follow
with
the
discussion,
to
hear
and
to
see
what
you're
doing,
and
I
don't
have
any
instruction
I
will
pass
to
yes.
F
Hey
everybody
yeah,
I
guess
I'm
also
with
jeff
and
thinking
you
know
what
is
possible
because
we
just
have
a
few
weeks.
I
do
think
sean
seeing
that
kind
of
like
paid
verse
not
paid
was
one
of
the
top
priorities
and
then
I
also
wanted
to
see
how
much
work
or,
if
possible,
since
we
spent
time
scoping
and
kind
of
doing
keywords
of
all
the
categories
or,
if
maybe
that
takes
a
lot
more
time
and
wouldn't
be
possible
by
hatch
time.
F
I
guess
still,
if
there's
any
possibility
to
just
look
a
little
bit
more
at
the
distribution,
because
that's
kind
of
one
of
the
main
things
even
outside
of
categories
like
looking
at
the
spread
of
distribution
and
then
yeah.
I
guess
just
trying
to
see
what's
possible
in
the
next
few
weeks
and
looking
at
like
our
hackmd
list,
but
we
had
started
working
on
the
categories.
F
So
I
guess
I'm
curious
what
we
can
do
with
that
and
wish.
I
could
help
more
with
the
data
side
yeah
and
I
will
I
also
was
a
few
minutes.
Late
went
for
a
grandma
visit,
so
I'll
just
pass
it
to
angela,
I'm
not
sure
who
had
a.
G
G
There
are
a
lot
of
steps
to
be
taken
and
yeah
we,
I
guess
everyone
was
motivated
to
do
this
exercise
to
find
out
if
this
has,
if
there's
a
something
that
is
really
important
for
the
hatch,
and
so
we
have
to
see
how
how
to
make
things
work
or
how
to
come
to
a
conclusion.
G
Compensation
from
the
hatch
impact
hours
to
tag
tokens
rather
on
the
long
term
impact
on
or
maybe
imbalance
on,
the
decision,
making,
power
and
distractions
a
little
bit
working
on
math
challenge
for
balancer
simulations
with
andrew,
and
I
think
you
I
can
hand
over
to
you,
octopus.
D
Yeah,
I
can
take
him,
we
can
yeah.
I
I'm
I'm
mostly
here
just
to
support
in
you
know,
questions
about
the
process
or
you
know
detailed
data
and
I'm
here
as
support,
but
we'll
probably
be
working
on
other
things,
while
you
guys
are,
are
chatting
and
and
doing
the
magic,
and
so
I
don't
have
any
direct
intentions
other
than
to
invite
you
all
to
the
params
party.
D
After
this
we're
having
the
the
great
debate
that
zeptimus
will
be,
recording
and
we'll
have
it'll
be
like
the
the
the
big
debate
where
we
talk
about
the
four
parameter.
Choices
sets
for
this
for
the
tec
economy,
which
actually
will
determine
greatly
the
importance
of
even
the
impact
hour
analysis,
because
if,
if
right
now,
there's
two
in
the
lead,
one
is
the
one
that
has
the
least
distribution
for
impact.
Our
people,
and
one
is
the
one
that
has
the
most
distribution
for
impact.
D
Our
people
so
pretty
interesting,
and
you
can
learn
more
about
it
at
right.
After
this
call
and
we'll
have
a
really
fun
debate
and
I'll
pass
it
to.
E
H
C
I
I
Yes,
thank
you
eduardo
and
yeah.
I'm
also.
My
intention,
for
this
call
is
mostly
just
learning
and
trying
to.
I
Get
deeper
into
the
various
corners
of
the
tc
and
also
like
I'm,
not
exactly
a
data
scientist,
but
I
have
manually
processed
mass
amounts
of
data
for
various
projects.
So
during
this
call,
if
there's
some
of
that
manual
grunt
work
that
needs
to
be
done.
You
can
definitely
tag
me
on
that
distractions.
I
I
will
say
well
at
the
moment
I'm
just
making
some
breakfast,
but
I
will
say
for
today
that
I
don't
have
too
many
distractions.
So
that's
good
and
I
think
so,
sulkin
have
you
be
gone.
C
Yep,
hey
my
attention
here
is
the
first
day
ever
joining.
So
I'm
just
here
to
listen
and
learn
in
that
showcase.
I
showed
sean
this
this
idea,
I
guess-
or
just
this
new
component
for
3.js
interesting
moment.
I
thought
it
would
be
an
interesting
way
to
display
some
of
the
words
or
the
phrases
that
have
gone
through
in
the
tc,
so
he
was
thinking
that
this
would
be
an
interesting
way
to
showcase
the
praise
data
just
by
everyone's
name
in
like
a
spherical
display.
A
C
J
I
don't
have
any
ability
to
type
at
this
moment,
but
I
would
like
to
come
up
with
some
clear
things
that
I
can
do
before
the
to
do:
asynchronously
before
windows,
wednesday's
meeting
so
sean
you
mentioned
dr
vek,
and
if
no
one
else
jumps
on
that
aspect
of
it,
I'm
willing
to
to
start
working
on
that
for
wednesday.
But
I
can't
do
it
today.
J
D
C
J
I'm
just
in
the
room,
I'm
hacking
on
the
the
mvp
dashboard
spec.
So
I'm
just
here
to
passively
listen
I'll
pass
it
to
whoever
hasn't
gone.
A
H
Hey,
I'm
just
here
to
support
and
watch
and
learn,
and
I
don't
really
have
any
distractions
hanging
out
in
my
office
and
I
don't
know
who
else
hasn't
gone.
D
A
I
think
it's
a
squad
call
but
sure
I'll
take
it
yeah,
it's
my
call
so
hosted
by
the
labs.
I
guess,
but
really
I
think,
jessica
and
angela
and
jeff
but
perfect.
So
that's
nice.
We've
got
all
this
from
everyone.
What
I'm
thinking
is
you
know
my
mind.
Just
goes
straight
to
the
data
science
stuff,
but
I'll
start
there,
octopus,
that's
great.
If
you
wanna,
I
just
had
a
feeling.
A
You
might
say
that
if
you
wanna
jump
on
the
dock
to
vec,
I
think
that's
nice
and
it's
falls
under
a
more
general
category.
That
is
like
encoding
the
data
basically
which
which
you
really
brought
up
last
time.
It's
a
really
good
point.
So
if
we
can
encode
all
of
the
columns
in
like
you've
mentioned
one
hot
encoding
for
the
categorical
variables.
A
A
You
know
unsupervised,
learning
for
clustering
and
t-sne
and
all
this
good
stuff.
So
I
think
the
encoding
is
one
of
the
biggest
things
that's
to
be
done,
and
the
doctevec
is
is
one
of
the
more
fun
aspects
of
that.
A
lot
of
the
other
stuff
will
be
like,
but
yeah,
the
one
hot
encoding.
Maybe
you
can
get
your
students
get
your
students
to
do
that,
but
yeah,
I
think
encoding.
A
The
data
set
opens
up
a
world
of
analysis
and
then
the
thing
I'm
going
to
jump
on
is
kind
of
a
separate
channel,
which
is
like
more
like
analytical
crunching
columns
against
each
other,
with.
A
A
Yeah,
so
I
think
that's
the
general
cluster,
like
there's
two
steps
to
this,
that
that
might
be
confused
and
one
is
the
embedding
of
so
a
lot
of
the
what
I've
been
thinking
about
and
what
we've
talked
about
is
really
focusing
on
this
column,
which
is
reason
for
dishing,
but
but
that's
really
an
inc
we're
using
the
we
can.
I
mean
we
can
use
unsupervised
learning
to
just
encode
this
column
by
using
this
thing
like
doctevec,
so
we're
already
getting
into
an
ai
process
which
isn't
even
to
get
results.
A
Really,
it's
intermediate
results
which
is
encoding
this
single
column,
but
we
can
get
so
we
can
turn
this
column,
which
is
right
now.
Each
row
is
a
string
of
text
or
yeah.
Each
row
is
a
string
of
text
like
a
sentence,
and
we
could
turn
this
into
a
like
five
columns
that
are
feature
vectors.
So
this
one
that
says
for
hosting
this
kicking
params
party
this
we
could
turn
into
like
five
columns
and
it
might
be
like
1.01
0.3,
negative
2.7.
A
So
it
would
you
just
imagine
this.
If
you
see
my
screen
here,
this
data
frame
would
just
get
like.
I
wish
you
could
see
my
hands
too,
but
this
data
frame
would
just
get
expanded.
I
mean
this
column
would
disappear
and
we'd
have
like
a
five
vector
of
real
numbers
that
expresses
the
spatial
embedding
of
this
column.
So
that's
that's
using
ai
to
just
do
an
encoding
of
a
single
column.
Now
we
have
all
the
other
columns
which
are
all
good
on
their
own,
like
I
said
we
can
actually
produce.
A
So
this
is
feature
engineering,
so
we're
at
the
feature
engineering
stage.
So,
for
example,
we
could
have
one
column
that
is
day
of
week
and
it
could
be
an
integer
from
you
know,
from
one
to
seven
or
or
simply
it's
a
categorical
variable
so
from
zero
to
six,
so
seven
options,
seven
categories
and
I
think
that'll
provide
some
rich
structure
in
the
data
and
then
these
names
should
be
one
hot,
encoded,
yeah
and
just
to
give
other
team
the
group
a
quick
synopsis
on
what
is
this
one
hot
encoding?
A
So
you
can
look
up.
You
know,
sk
learn.
So
one
hot
encoding,
sk
learn
is
like
the
standard
machine
learning
processing
toolkit
for
python.
So
you
get
this
one
hot
encoder,
so
you
encode
it
categorical
feature
as
one
hot
numeric
array.
So
I
don't
think
I'll
go
deep
into
this,
but
definitely
good
reading
for
any
anyone.
A
You
know
curious
I'll,
just
pop
this
in
the
labs
channel
and
it's
a
way
of
saying
instead
of
like
because
you
could
encode
this
column
numerically,
so
you
could
say:
okay,
zeptimus
is
the
number
one
atticus
is
two
luke
weber
is
three
crysis
is
four
sebastian
is
five,
and
so
there
you
get
a
numerical
encoding,
which
everything
needs
to
be
numerical
in
the
end,
when
you
pass
it
into
a
machine
learning
algorithm,
but
what
this
does
is
it
introduces
bias
because
the
algorithm
is
going
to
start
trying
to
assign
meaning
to
the
magnitude
of
the
encoding.
A
So
if
you
say,
zeptimus
is
number
one
and
chris
is
number
seven,
then
in
some
ways
that
means
chris,
it
means
zeptimus
times
seven
equals
chris,
and
so
you
get
this
weird
kind
of
algebra
structure
that
doesn't
actually
have
meaning.
So
it's
better
to
not
just
do
a
direct
numerical
encoding
with
like
integers
it's.
What
you
need
to
do
is
for
each
cat
each
option
for
the
category.
A
A
The
bad
part
is
that
it
kind
of
explodes
your
column,
your
columns,
because
for
now
for
every
categorical
column
that
you
have
you're
getting
a
new
column
for
every
value
in
the
volume
column,
so
you
can
get
this
it's
like
you
could
potentially
get
like
n
squared
or
you
know
you
can
do
the
math
on
this,
so
that,
if
m
is
the
number
of
categories
then
and
and
or
you
the
average
number
of
options
in
a
categorical
column,
and
you
have
n
categorical
columns,
then
you
now
have
n
times
m
columns
instead
of
just
n,
so
it's
kind
of
a
square
growth,
so
there
can
be
trade-offs,
but
it's
the
default.
A
Usually
when
you
have
a
categorical
column,
you
encode
it
using
one
hot.
So
it's
probably
what
we'll
do?
I
think
it'll
work
fine
and
then
even
like
this
room,
the
room
is
categorical
and
then
even
these
norm,
columns
see
this
is
where
some
domain
expertise-
and
this
is
what
octopus
said
when
it
comes
to
the
encoding.
A
We
might
need
community
people
who
are
deep
in
the
community,
because
I
know
personally,
I
haven't
yet
built
the
intuition
around
these
norm,
columns
and
how
we
can
how
they
should
look
in
the
analysis
and
how
we
can
really
utilize
them.
Maybe
we
just
normalize
these
numbers,
so
we
squash
them
all
between
negative
one
and
or
just
between
zero
and
one.
I
don't
know
if
that
would
maintain
the
integrity
of
the
information
that's
held
within
them
or
we
might
be
able
to
use
them.
A
Analytically,
like
directly
like
we
take
these
number
this
number
and
we
divide
it
by
the
amount
of
praise
that
was
given
something
along
those
lines.
But
this
is
really
what
octopus
said.
The
exploratory
data
analysis
step.
I
mean,
I
think
it
starts
with
the
distributions
right.
So
this
is
really
you
got
to
walk
before
you
can
run.
So
if
we
render
a
distribution
of
this
v1
norm,
then
at
least
we'll
have
a
sense
of
what
these
numbers
look
like
like.
What's
the
domain?
A
What's
the
distribution,
so
I
really
do
think
that's
step
one
but
yeah.
That
was
just
kind
of
a
lot
of
an
introduction
for
anyone
who's
interested
in
some
of
the
deeper
stuff
I
mean
the
concept
was
that
we're
using
ai
just
column,
but
all
the
columns
can
be
encoded
and
then
and
then
we
can
use
ai
across
the
whole
data.
Basically,.
F
A
J
D
Yes,
it's
I
it's
ih
per
praise,
the
name
of
the
row,
I'm
sorry
the
name
of
the
column
for
the
header
of
the
poem.
A
Max
is
48.
48
hours
on
a
single
praise.
Interesting,
that's
a
big
one,
probably
sem,
but
yeah
this.
I.
D
Assume
that
was
actually
simon
de
la
riviere
for
inventing
bonding
curves.
D
D
F
F
D
You
know
you
say
the
distribution
is
nuts
but
th.
This
is
a.
It
depends
on
how
you're
looking
at
it.
If
it
was
an
airdrop,
the
distribution
would
be
really
bad,
but
if
it's
like,
you
know
for
work
done,
given
certain
parameters
or
whatever,
then
it
it
like.
If
this
was
a
distribution
of
mining
rewards,
you
know
it
wouldn't
be.
Like
I
mean
it's
not,
but
like
you
know
it's
it's,
it's
like.
I
think
the
output
is
okay.
Here's
the
data,
here's
I
liked
your
category
idea
like
is
this?
D
How
we
want
to
now
that
we
have
this
data
like?
Does
this
distribution
make
sense
given
because,
like
just
saying
the
distribution
is
bad?
I
don't
know
that.
That's
like
enough.
We
need
to
like
justify
why
it
got
out
of
hand
or
how
it
got
out
of
hand
in
a
way
that
the
community
may
not
appreciate
right.
A
That's
why
I
really,
I
think,
the
best
way
this
isn't
just
all
for
fun.
This
force
directed
graph.
I
think
it's
gonna
give
a
very
accurate
picture,
because
we
can
actually
weight
the
edges
by
the
strength
of
the
praise,
so
we
can
see
not
just
like
what
were
the
clusters
of
sort
of
instances
of
praise.
It's
not
just
like.
Where
were
people
praising
each
other
a
lot,
but
we
can
actually
have
it
fully
weighted
in
its
structure
by
the
amount
of
praise
given.
A
So
we
can
see
the
density
of
praise,
actual
praise
distribution
from
something
like
this,
because
each
of
these
edges
you
can
parameterize
it
with
yeah
strength
and
length,
and
so,
if
we
parameterize
each
each
edge
in
a
force
directed
graph
with
the
actual
magnitude
of
the
praise
that
was
dished,
we
can
see
the
structure
and
then
I
think
key
is
what
what
you
definitely
see
jessica
and
a
lot
of
people
see
here
is
the
the
grouping
is
necessary.
Absolutely
it's
like.
A
So
if
we
can
get
deep,
if
we
can
use
ai
to
all
of
these
columns
and
categorize,
I
think
it's
an
essential
step.
I
think
it's
like
okay
by
the
by
the
time
we
out
have
the
outputs
for
this
project,
and
maybe
I
think,
a
week
from
next
tuesday,
so
we
will
have
a
categorization
phase
and
then
we
can
see
the
the
categories,
along
with
the
distribution
of
amounts
like
just
in
a
simple
sort
of
histogram
style
like
like
yeah,
a
dist.
A
What's
the
total
praise
given
and
then
what
is
that
as
a
percentage
of
all
praise
given
so
sort
of
this
table
and
and
but
and
then
it's
this
categorization,
the
bucketing,
I
think
we
can
get
out
a
first
version.
However,
we
do
that
and
it's
a
very
iterative
process,
so
we
can
see
what
those
categories
those
buckets
look
like
and
we
can
improve,
improve,
even
use
like
we
might
start
with
one
algorithm
and
then
use
completely
different
algorithm.
A
F
Yeah,
that
was
my
main
question
because
I
was
kind
of
like
you
know.
It
sounds
like
it's
like
you
think.
Maybe
you
can
do
it
without
that,
but
I
feel
like
we
have
people
here
and
I
know
for
me.
I
can't
really
help
on
the
data
side.
So
is
it
useful
and
worthwhile
if
we
go
ahead,
I
mean
in
a
praise:
quan
we've
we've
quanted
like
last
week,
2000
praises,
so
I
feel
like.
If
a
couple
of
us
did
this
kind
of
like
categorization
or
bucketing,
it
would
not
take
very
long.
C
J
F
So
then,
if
we're
gonna
do
this
and
it
is
really
useful
it
best
to
just
take
like
one
month's
example
or
one
praise
round
example,
or
should
we
draw
like
100
randomly
from
each
one
and
then
build
a
spreadsheet,
and
then
we
were
discussing
last
time?
Is
it
better
to
key
by
number
or
words
and
octopus,
you
kind
of
said
words
and
then
sean?
You
said,
maybe
actually
numbers
in
case
there's
any
like
typos.
A
Yeah
well
kind
of
like
what
you
have
here,
which
and
then
you
have
the
description
or
you
have.
Let's
see,
okay,
a
category
sort
of
the
category
title,
so
you
guys
already
have
both
here,
which
is
really
nice.
The
category
and
the
description
and
keywords
like
this
is
pretty
awesome,
I'm
wondering
so.
A
I'm
just
gonna
talk
this
through.
I
don't
know
the
best
way
to
do
it,
I'm
curious
what
octopus
will
think
and
everything,
but
the
idea
that
I'm
having
is
basically,
I
think
we
should
take
this
whole
data
set
and
we
should
randomly
sample
it
so
like
basically
shuffle
it
and
take.
What
do
you
think
two
like?
Maybe
we
can
is
500
too
many.
J
C
C
A
F
So
can
you
yeah
if
500
that's
fine,
like
I
can
do
that
myself
in
a
half
an
hour,
probably
but
there's
only
9,
000
or
whatever
total.
So
we
could
start
with
that
and
see
how
far
and
can
you
put
it
in
the
same
spreadsheet
as
the
categories.
C
F
Then
show
like
how
you'd
want
me
to
label
it
so
that
and
then,
if
anybody
else
here
wants
to
like
support,
we
can
just
take
like
I'll.
Take
one
through
100
start
somebody
take
200
through
300
and
just
see
if
we
can
crash
them
out
and
then,
if
there
are
questions
about
categories
we
can
decide
together
in
case
a
question
comes
up
I'll,
just
share
this
category
sheet
in
the
pram's
channel.
F
F
G
F
Yeah,
it's
mostly
straightforward.
There
might
be
a
few
like
I
don't
know.
I
think
it's
pretty
straightforward,
but
we
can
see,
as
it
comes
up
yeah.
G
F
The
other
thing
I'm
kind
of
looking
forward
to
see
too,
is
like
how
this
like
participation
like
attending
calls
and
meetings
and
like
retweeting
and
stuff
stacks
up
against
some
of
the
other
work
like.
If
you
pull
it
out
as
a
category,
is
it
weighted
a
lot
higher
than
some
of
the
other
work,
because
what
is
seen
it's,
it's
very
like
visible.
F
So
weird
paradoxes
like
I
feel,
like
the
praise
system,
is
really
great
at
quantifying
this
invisible
work
but
weirdly
it
almost
values.
This
normally
invisible
work
more
than
the
visible
work
like
in
a
in
a
skewed
way.
So
it's
just
it's.
C
B
And
this
is
one
of
the
points
I
made
is
that
you
know
the
praise
system.
I
think,
is
a
a
composable
mechanism.
You
know
when
we
have
like
a
real
world
payment
structure,
for
example
like
salary
or
something
that
you
know.
We
don't
have
this
ability
to
recognize
the
unseen
labor
above
and
beyond
that.
So
I
think
the
prey
system
is
a
really
interesting
posable
mechanism
and
I'm
curious
how
we
choose
to
move
forward.
You
know
whether
integrating
I
feel
like
source
crest
sees
part
of
the
landscape.
B
B
So
I'm
interested
how
we
can
leverage
multiple
tools,
including
you
know,
the
prey
system,
source,
cred,
a
regular
salary
system
and
potentially
even
more
to
view
as
much
of
this
landscape
as
the
community
finds
relevant,
and
I
think
this
discussion
came
up.
You
know
the
praise
system
is
really
good
at
accounting
for
certain
types
of
labor,
but
not
so
great
at
other
types
of
labor.
So,
just
looking
at
you
know
how
certain
actions
may
have
been
undervalued
and
how,
as
a
community,
we
want
to
make
sure
those
actions
are
recognized
and
valued.
A
Yeah,
the
composability
is
super
exciting
when
you
think
about
this
yeah.
This
is
just
one
piece
and
there's
like
100
innovations
to
come
on
how
to
do
and
there's
a
lot.
We
have
a
lot
of
cons
mixed
together
here,
like
governance,
tokens
and
salary
distributions,
compensations
and
all
sorts
of
stuff.
I
think
through
really
really
deep
analysis
over
time.
A
We
can
actually
splice
those
things
out
and
decouple
these
concepts
and
have
like
a
very
orthogonal
mechanism
toolkit
for
having
higher,
very
high
precision
like
on
rewarding
someone,
monetarily
versus
having
someone
earn,
governance
influence
and-
and
maybe
it's
it's-
it's
probably
okay,
and
maybe
even
good-
to
have
these
things
combined
to
some
extent
and
to
have
overlap
but
to
have
full
control
over
that
and
insights.
Through
the
analysis,
I
think
yeah
the
composability
is
to
me.
Maybe
the
most
interesting
aspect
of
all
of
this
so
well
said:
jeff.
F
Oh,
that's
not
me
who
does
that
great
griff
is
a
spreadsheet
wizard.
A
C
A
A
F
A
G
A
C
A
Just
the
ability
to
so
like
cad
cat
cad
cat
is
a
subcategory
so
when
to
be
nice
to
all
praise
that
was
related
to
cad
cad
verse
token
spice
first.
B
There
a
way
we
can
do
this
as
a
first
pass,
but
and
and
then
decide
if
we
need
to
get
more
granular
from
there,
like
I'm
just
looking
at
some
of
these
like
token
spice,
if
it's
received
any
praise,
maybe
one
or
two
token
cab.
If
that
and
I'm
just
curious,
how
granular,
like
I
mean
even
19
categories,
seems
like
a
lot
already
before
we
break
into
subcategories.
A
It's
a
good
point:
jeff,
there's
nothing
stopping
us
to
later
add
a
subcategory
column
and
yep,
and
then
everyone
could
just
yeah.
We
could
go
through
that
and
add
all
this.
If
necessary,
keywords
are
pretty
neat.
I
think
I
would
like
to
have
the
keyword
data
set.
A
G
A
Yep
good
point
so
like
from
the
data
like
a
bottom
up
from
the
data
keywords
associated
with
a
category,
I
feel
like
having
these
two
label
columns
adds
flexibility
and
like
cross
reference
ability,
so
we
might
have.
We
have
largely
working
groups
on
the
left
and
on
the
right.
We
might
have
something
like
cad
cat.
We
might
see
that
cad
comes
up
in
or
github
or
transparency
or
they're.
I
think
just
this
idea
of
having
two
categorical
columns
allows
for
like
cross-reference
ability.
F
J
F
F
G
G
H
F
C
I
G
E
C
I
A
F
So
I've
got
this
sheet
here
and
angela,
and
I
and
a
few
others
built
these
categories,
so
we
put
a
number
and
then
we
put
a
category
so
I'm
going
to
go
to
the
praise
sheet
and
to
the
best
of
my
ability,
I'm
going
to
look
like
for
this
one
says,
always
carry
caring
for
our
bought
and
valued
community
members.
Okay,
that's
a
little
mixed
actually
because
kind
of
like
doing
bot
work
and
also
it's
personal
praise.
So
sean
actually
is
a
good
point.
Should
I,
if
these
things
are
two
things,
what
do
we
do?
F
F
F
Okay,
so
in
this
case,
we've
got
somebody
who's
like
doing
bot
work,
so
that's
like
tech,
dev
and
then
we
have
like
caring
about
community
members
and
I
think,
that's
kind
of
like
what
we
could
call
that
personal
praise
or
we
can
call
that
participation
or
community
building.
Is
there
a
community
building.
C
F
E
B
Just
need
to
make
you
know
pick
one
category,
I
I
have
a
feeling
like
trying
to
figure
out
if
this
is
like
within
one
or
within
two
I'm
just
trying
to
think
like
how
do
we
do
this
within
a
amount
of
time?
That's
reasonable
and
I
think
it's
like
just
let's
put
it
in
a
category
and
move
to
the
next
one,
because
otherwise
this
is
going
to
take
us
weeks
just
to
label
this.
G
Yeah,
but
actually
I'm
missing
a
category
that
is
about
this
community
thing.
We
have
com
communications
and
we
have
cultural
build,
but.
F
F
C
G
F
Okay,
I'm
just
gonna.
I
can't
even
see
some
of
this
one
that
I'm
gonna
fix
it.
Okay,
so
we
have
comms
ultra
build.
F
G
F
Okay,
so,
like
a
tec,
community
building,
yeah
that'll
be
great.
I
think
that'll
hopefully
solve
that
solve
for
that.
Okay,
so
we're
doing
one
category
only
to
the
best
for
our
ability.
So
this
one
is
like
always
caring
for
our
spot
and
community
members.
So
I'm
I
mean
I'm
gonna
call
that
infrastructure
work
because
he's
taking
care
of
a
bot.
F
F
F
C
G
A
G
They
need
to
be,
since
we
have
this
overlap
of
very
similar
tasks
to
in
in
both
areas,
I
think,
having
the
words
the
name.
Category
name
only
can
be
confusing,
and
then,
if
you
have
a
typo,
then
yeah
you
have
a
new
category.
So
this
is.
A
G
A
F
I
I
All
right
I'll
be
jumping
right
on
that
50.
So,
okay
off
my
turf.
F
F
F
F
G
G
F
Oh,
yes,
sorry,
I
already
started.
Oh
one
more
thing:
what
did
we
say
about
retweets
and
all
that
we
also
consider
that
participation.
H
No,
the
I
I
think
the
retweets
are
didn't.
We
just
say
that
we
want
to
separate
out
the
retweets
and
the
just
the
meeting
attendance
from
the
other
things.
Okay,
I
don't
know
I
mean
I
understand,
but
I
think
retweets
is
more
participation.
F
F
F
E
G
F
C
F
E
Well,
I
think
that
gravity
is
a
working
group
and
it
can
be
like
a
bucket,
but
I
I
am
feeling
like
a
little
bit
concerned
about
about
like
making
these
buckets
because,
like
I
remember
that
for
the
first
rounds
we
had
we
tried
to
do
this
tier
tiering
system,
and
I
think
that
it's
difficult
yeah
to
to
to
there
are
gonna,
be
a
lot
of
places
that
are
going
to
be
like
in
the
middle
between
one
bucket
or
another.
F
Yeah
going
through
this,
like,
I
think,
we've
sorted
out
some
of
the
main
ones,
but
you
know
juan
if
you
want
to
kind
of
look
through
and
just
like,
have
a
look
at
how
we're
bucketing,
but
I
feel
like
it's
pretty
clear
thus
far,
except
for
the
few
flags
that
we've
raised
and
fixed
then
as
we
go,
we
can
just
see
if
there's
any
other.
If
anybody
has
like
any
concerns,
but
I
feel
like
it's,
I
don't
think
it's
too
many
on
the.
F
E
F
Yeah,
the
praise
is
quantified
like
so
the
double
shouldn't
matter,
but
I
see
that
as
personal
praise,
because
it's
like
meeting
someone
and
then
a
new
project,
not
that
the
art
project,
to
my
knowledge,
isn't
tec
or
te
unless
there's
some
part
of
it.
That's
going
to
be
open
source,
but
it
seems
like
a
personal
project
like
if
somebody
gave
me
praise
for.
F
E
That
makes
sense
I
have
gave
given
several
praises
to
to
mateo
for
collaborating
with
the
ed
colombia
community.
Would
that
be
personal
as
well.
F
C
F
Talks
and
events-
and
I
think
it's
expanding
like
the
te
community
and
also
to
ec-
you
can
make
a
case
for
either,
but
this
this
won't
affect
waiting.
It's
just
like
how
we
are
going
to
cluster
and
look
at
the
difference
between
some
of
the
work
and
what
was
weighted
more
heavily
so
like.
F
We
would
see
the
difference
between
like
stewards,
work
for
the
tec
and
then
like
people
retweeting,
let's
say
but
yeah
that
work
to
me
would
be
te
community
building
or
ecosystem
development,
in
my
opinion,
but
if
other
people
have
a
different
idea
and
yeah,
I
do
think
it's
helpful.
We
sense
make
together
on
this,
because
then
everybody
will
feel
better
feel
good,
that
it's
like
makes
sense.
C
F
E
Or
also
culture
build
soft
gob.
F
F
J
I
F
I
Well,
it's
like
participation,
that's
what
wasn't
really
my
question.
I
got.
That's
fine
can
say
it's
participation,
but
I
just
wasn't
sure
if
I
should
try
and
because,
like
I
am
I
get
caught
in
the
details
like
that.
J
I
F
And
but
yeah
you
shouldn't
alter
you
shouldn't,
alter
any
of
the
data
there
better
you
flag
it,
and
then
we
can
look
at
it
together.
I
think
yeah.
The
main
thing
is,
if
you
know
its
participation
like
for
attending
a
meeting
or
whatever
you
can
throw
the
code
on
there
still
and
then
we
can
look
at
often
people.
F
E
G
F
C
F
G
F
C
F
F
E
Interesting,
what
about
the
comps
that
are
done
for
the
common
swarm,
for
example,
like
3,
plays
1.5
hours
of
work,
the
weekly
date
to
the
prep
for
tuesdays
m.a,
and
it
was
for
the
common
swarm.
It's
not
a
specific
guide
in
here,
but
I
know
it's
from
commons
one.
So
I
don't
know
how
should
we
approach
this
data.
F
It
was
oh
comms
work
for
the
common
swarm.
I
would
still
put
it
under
common
swarm
because
yeah
they
did
an
ama,
unlike,
I
think,
which
might
call
it
celeste,
so
they
were
just
like
sharing
about
their
common
swarm
work.
So
I
would
say
common.
C
F
F
F
F
I
F
I,
let's
see.
F
F
Dapp
and
things
like
this
yeah
and
then
the
technical
infrastructure
is
like.
There
are
other
people
that
are
like
handling
bots
and
doing
modeling
work,
but
we
could
merge
it
into
one
stream.
I
I
I
was
focused
on
that.
I
didn't
hear.
Oh,
I
was
just
saying:
what's
the
difference,
tech,
technical
infrastructure
and
common
swarm
category,
or
should
we
say
common
swarm
is
specifically
for
the
commons
related
stuff,
where
technical.
C
H
C
D
D
I
Okay,
well,
that
that
was
just
my
evidence
of
my
naive
understanding
of
what
the
commons
forum
was,
because
I
just
took
it
to
me
like
that.
That's
the
group
that
does
all
the
technical
stuff.
D
I
F
F
H
F
C
E
I
I
am,
I
am
not
taking
like
one
range.
I
am
looking
for
the
ones
that
I
can
like
easily
detect.
F
Okay,
so
I
just
found
one
as
an
example
of
technical
infrastructure
for
anybody
once
so.
There
was
a
praise
for
bot
being
deployed,
so
I'm
I'm
giving
that
a
technical
infrastructure,
because
it's
not
necessarily
a
common
storm.
It's
just
like
tech,
a.
F
F
And
same
like
transparency,
writing
a
blog
of
trend
about
transparency
like
it
could
becomes,
but
to
me
really:
it's
transparency,
so
I'm
gonna
put
transparency,
work.
F
C
C
C
D
I
can't
see
the
spreadsheet
but
yeah
I
would
also
go
with
with
zepp
on
the
prams.
It's
definitely
not
swarm.
Thank
you.
C
C
C
C
C
F
F
F
I
B
There
are
a
few
people
that
are
taking
the
taking
one
thing
and
going
across
the
whole
list.
At
least
that's
that's
one
thing
I've
been
doing
just.
I
found
it
easier
to
pick
out
when
I'm
looking
for
one
thing,
so
you
may
see
a
couple
of
random
ones
scattered
through,
but
don't
don't
mind
those
or
confirm
them
to
make
sure
you
agree.
I
guess.
B
C
B
C
So
we
have
the
the
one
that
used
to
belong
to
the
boarding
working
group.
That
became
the
hatch
outreach
working
group.
So
shall
we
label
it
on.
F
Steel
yeah
like,
if
you
look
at
the
category
it
says
catch
outreach
onboarding,
it's
like
we
have
both.
F
F
C
H
F
F
Oh
celeste,
sorry
yeah,
I
guess
it's
common
swarm
or
it
could
he
be
now
it's
common
swarm.
There's
nothing
t
frameworks
and
tools.
H
E
F
Let's
I'm
scrolling
there's
this
girl
and
see
if
you
see
any
blanks
like,
for
example,
I
just
found
two
blank:
three,
two,
five
and
three
two
six
somebody
missed
or
wasn't
sure
how
to
do
those,
maybe
just
scroll
and
see
where
it's
blank
looks
like
we're
doing
pretty
good
looks
like
we're
almost
done
so
just
scroll
for
blanks
and
I'll
scroll
with
you,
because
I
also
finished
my
section,
so
I'm
just
gonna
scroll
and
see
if
anybody
missed
or
was
like,
maybe
skip
them
because
they
weren't
sure,
okay,
yeah,
that's.
H
H
E
H
I
think
371
have
question
marks.
I
feel
like
ecosystem
development
might
be
a
good
fit
for
those
okay
which
number
he
is
oh
367,.
H
F
F
F
H
H
Common
swarm,
because
that's
what
is
grif
is
he
still
listening.
H
It's
already
illegal
yeah,
so
on
number
three
six,
oh
you
don't
have
the
spreadsheet.
So
sam
got
a
complicated
phrase
that
covers
several
working
groups.
Should
we
defaulted
to
common
swarm.
D
Yeah,
that
was
a
tough
one,
yeah
yeah,
I
guess
default.
You
just
pick
one
of
the
three
right.
This
is
what
octopus
said.
I'm
sorry
right.
D
D
C
I
So
I
just
want
to
say
clear:
somebody
was
mentioning
room
like
participating
in
the
forum
as
just
participation,
but
to
me
that's
more
like
comms
work,
because
working
responding.
Writing
in
the
forums
is
a
lot
bigger
than
a
retweet
or
whatever
just
my
opinion.
H
I
C
Yeah,
I
think
that
one
it's
on
boarding
actually,
because
it's
this
katie
is
taking
care
of
morning.
I
think
those
meaning.