►
From YouTube: TechTalk: DApp Curation by Andy Tudhope
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
So,
as
you
all
know,
the
best
place
to
start
a
presentation
is
at
the
cute
Doge,
but
we're
now
on
the
main
nets.
And
so
one
of
the
next
big
problems
that
we
have
to
face
is
figuring
out.
Not
only
what
optimized
for
status
might
mean
for
different
DFS,
but
also
how
they
actually
rank
in
our
store
and
how
users
can
discover
them
in
different
contexts
and
do
that
easily
and
intuitively,
without
needing
to
go
through
a
whole
educational
rigmarole
or
an
understanding.
A
What
it
is
that
they're
doing
so
in
order
to
sort
of
answer
this
I
think
the
best
place
is
to
start
with
some
history,
and
the
best
questions
ask
is
what
made
PageRank
famous,
which
is
the
algorithm
that
Google
uses
to
index
their
searches,
and
you
know
I
think
that
the
core
insight
here
is
a
quote
from
Marshall
McLuhan.
Who
once
said
that
the
medium
is
the
message
and
Larry
Page
and
Sergey.
A
Brin
certainly
recognized
this
in
the
early
2000s
when
they
came
to
the
idea
that
the
central
problem
facing
the
web
at
that
time
was
not
getting
to
was
not
getting
documents
to
display,
as
we
might
expect
them
to
in
the
real
world,
which
is
what
everybody
else
was
trying
to
do
at
the
time,
but
rather
organizing
the
fire
hose
of
information
that
the
web
was
generating
through
this
sort
of
massive
collaboration,
creative
explosion.
That
was
going
on
and
that's
what
sort
of
led
PageRank
and
you
can
see
a
really
simplified
version
of
that.
A
It's
obviously
much
more
complicated
these
days.
But
the
basic
idea
is
that
any
page
has
sort
of
incoming
and
outgoing
links
and
number
of
incoming
links
is
what's
most
important,
are
drinks,
but
the
weights
that
it
is
given
by
the
algorithm
and
therefore
how
highly
a
drinks
in
your
search.
Obviously,
that
is
manipulated,
though,
by
advertising
in
particular,
and
the
other
big
point
is
that
you
don't
get
to
see
the
inner
workings
of
pay
drink.
A
You
just
sort
of
have
to
believe
Google
and
trust
them
to
do
this
weighting
correctly
and
that's
kind
of
a
central
problem.
Now,
isn't
it
that
it's
no
longer
a
question
of
just
organizing
information,
that's
about
who
gets
to
organize
the
information,
how
they
presented
to
the
users
and
what
they
choose
to
leave
us.
So
it's
really
about
what
personalization
means
in
a
world
of
targeted
ads
political
manipulation
and
what
is
essentially
a
total
lack
of
privacy.
A
You
know
the
the
I
think
a
lot
of
people
in
status,
particularly
through
some
of
the
discussions
around
analytics
and
Mixpanel,
have
come
to
the
recognition.
That's
data,
tracking
and
understanding
how
product
is
used
is
not
necessarily
a
bad
thing
at
all.
It's
in
facts,
you
know
incredibly
beneficial
to
or
making
a
much
more
user-friendly
experience
in
general
and
understanding
where
it
is
that
we're
ready
losing
people
in
the
journey
to
web
3.
A
So
how
do
we
solve
this
problem?
Well,
you
know.
Sirkin,
curated.
Registries
are
one
possible
option.
I
have
one.
It's
include
that
neat
picture,
because
I
think
it's
pretty
awesome
but
you'll
see
here.
This
is
the
one
of
the
early
beta
versions
of
the
ad
chain,
so
I
can
curated
registry
for
advertisers
and
the
content
that
they
published,
and
you
know,
I'm
a
massive
fan
of
Matt
golden
and
the
guys
at
edge
and
I
think
that
they're
pretty
intelligent
people.
A
A
A
A
With
that
token,
you
know:
there's
gonna,
be
this
endless
accumulation
of
tokens
in
my
wallet
that
I
don't
really
know
what
to
do
with
and
that's
quite
a
concern
when
we
think
about
mass
adoption
of
systems
like
ethereum
as
law.
So
that
suggests
that
it's
moving
towards
more
accurate
consumer
cribs
economics
which
sort
of
based
around
this
primal
assertion.
That's
the
whole
points
of
mechanism.
Design
of
cribs
economics
is
to
make
sure
that
there's
no
single
central
point
of
failure,
including
the
community
and
I,
think
that
this
is
a
really
really
key
points
right.
A
A
Unlike
the
Silk
Road,
these
systems
tend
to
be
very,
very
easy
to
manipulate
it's
very
easy
to
buy
Twitter
followers
or
to
buy
views
for
your
YouTube
videos
or
various
other
things
that
can
then
be
used
to
influence
these
advertising
metrics,
because,
again,
the
whole
system
is
trying
to
optimize
for
the
wrong
thing,
which
is
advertising
revenue
and
user
information,
rather
than
collaboration
and
creativity
and
the
sort
of
move
toward
a
collective
kind
of
knowledge.
That's
actually
curated
and
interesting
to
a
large
community
of
people.
A
A
Pure
price
discovery
mechanisms
are
open
to
manipulation
by
people
who
have
lots
of
money,
obviously
right,
so
they
work
very,
very
well,
but
in
already
unequal
societies
they
are
open
to
various
kinds
of
manipulation,
which
is
precisely
why
we
have
regulation
and
anti
monopoly
laws
in
place.
Those
things
are
useful.
There's
a
good
history
behind
them
that
you
can
go
in
and
look
into.
But
the
point
here
is
that
we
don't
want
human
regulation
right.
We
want
maths,
that's
the
core
insight
of
Bitcoin.
A
We
don't
necessarily
need
a
whole
bunch
of
human
bodies
and
generally
generally
agreed
accounting
principles
and
King
four
and
whatever
else
this
stuff
is.
We
can
just
use
maths
and
cryptography
to
regulate
the
systems
that
we
use
and
allow
for
this
trustus
execution
of
arbitrary
code,
which
is
the
feature
of
ethereum.
A
So
we'll
make
this
rule,
that's
whoever
stakes
the
most
SNT
ranks
the
highest,
with
the
one
caveat
that
the
more
SMT
you
stake
as
a
developer,
so
rank
high
D
the
easier
it
is
for
the
community
to
influence
that
position,
and
in
order
to
that,
in
order
to
achieve
that,
we're
gonna
have
to
use
the
site.
This
one
idea
from
TCR
is
not
the
whole
commits
just
one
idea,
which
is
bonded.
Curves
I'll
come
back
to
the
slide,
but
this
is
basically
what
I'm
talking
about
right.
So
you
can
see
the
x-axis
down.
A
So
what
we
can
then
do
in
terms
of
constructing
these
graphs
and
actually
understanding
how
we
are
going
to
achieve
those
four
features
that
I
said
with
key
for
any
user.
The
first
thing
that
we
have
to
do
is
map
the
costs,
the
cost
that
it
actually
has
to
mince
these
votes
in
order
to
add
votes
or
download
the
app
in
our
store.
You
can
see
that
that's
also
a
decreasing
exponential
function.
It's
just
the
inverse
of
the
number
of
tokens
that
are
printed,
I
mean
a
number
of
votes
and.
A
Like
here's,
some
extra
detail
in
case
you're
struggling
to
follow
along
so
for
every
SNT
you
stake
to
rank
in
the
store.
A
certain
number
of
votes
are
minted,
the
more
SNT
you
stake,
the
more
votes
on
minted
using
that
exponential
curve
that
I
just
showed
you.
Those
votes
can
then
be
bought
by
anyone
and
used
to
either
up
votes
or
down
votes
your
gear.
It
costs
people
to
do
to
buy
those
votes
and
that
cost
goes
directly
back
to
the
developer.
But
that's
a
part
that
I've
come
to
in
another
slide.
A
Importantly,
the
more
votes
that
have
already
been
cast
the
more
expensive
it
is
demand
further
votes.
When
someone
down
votes
your
yep,
that
does
subtract
value
from
your
effective
balance,
which
is
the
thing
that
is
used
to
rank
the
apps
in
the
store.
So
this
is
why
this
idea
to
me
is
so
exciting
because
you
can
imagine
having
a
separate
tab
in
status.
A
So
this
is
beginning
to
sort
of
map
out
some
of
some
of
these
subtleties
right.
Instead
of
just
having
the
red
like
the
red
cost
curve,
and
the
blue
number
of
votes
demand
curves,
we
then
say,
unfortunately,
not
all
of
the
SMT
that
is
staked
by
the
app
developers
to
rank
in
the
store
can
actually
be
available
to
mint
votes
with
because
it
needs
to
another
kind
of
attack.
So
we
have
to
say
that
only
a
certain
percentage
of
the
S&T
that
is
staked
is
actually
available
to
men.
A
Folks
with,
and
the
problem
is,
how
do
we
specify
that
centage?
How
do
we
know
what
what's
best
right,
because
the
higher
the
percentage
is
like
the
more
of
the
SMT?
That
is
stakes
that
can
be
instrument?
Votes
with
the
slower
SNT
will
accumulate
in
the
registry,
and
the
total
SMT
in
in
the
dev
store
is
like
a
really
good
proxy
for
the
overall
crypto
economic
security
of
the
entire
store.
A
It
doesn't
touch
the
status,
it
doesn't
go
into
contracts
if
you
buy
votes
on
a
certain
DF
by
either
clicking
uploads
or
downloads
that
costs.
In
essence,
he
goes
directly
back,
so
we've
we've
done
that,
and
you
can
see
that
the
new
curves
here,
the
orange
ones
so
they're
slightly
less
SNC,
meant
to
I
mean
slightly
less
votes,
minted
Basin
C,
which
is
what
you'd
expect,
because
we
only
using
a
certain
percentage,
and
that
means
that
the
cost
is
slightly
higher,
that
Green
Line,
which
is
great
we're
getting
closer.
A
Now
we
need
to
implement
the
last
little
subtlety,
which
was
that
the
more
votes
that
have
been
cast,
the
more
expensive
that
is
to
mint
further
votes
right,
because,
ideally
speaking,
what
you'd
once
is
the
closer
that
your
vote
gets
to
actually
moving
the
d
up
in
the
stores
actually
having
a
certain
effect
on
the
actual
rankings,
the
more
expensive.
It
should
be
two
votes,
but
there's
no
easy
way
to
measure
that
accurately.
A
So
because
we're
using
these
smooth
exponential
curves,
we
can
apply
the
same
thinking
and
the
more
votes
that
have
been
cast
already,
as
a
percentage
also
should
affect
how
expensive
it
is
to
mint
further
votes.
So
it
makes
it
more
expensive
and
you
can
see
the
final
lines
here-
the
purple
one
and
the
light
blue
one.
A
So
so
that
seems
to
have
got
us
quite
close
to
what
we
need
for
the
system
to
work,
and
this
is
one
of
the
key
points
right
that
even
some
even
better,
whatever
it
costs
to
uses
the
community
to
vote,
goes
directly
back
to
the
to
the
developer.
So
it
means
that
back
even
if
we
starts
a
DAP
Wars
situation
or
developers
become
victims
of
trolling
or
corporate
warfare
or
targeted
community
campaigns,
they
still
receive
the
value
that
they
stake
back
as
they
move
down
the
rankings
right.
A
I.
Think
that
the
key
points
here
right
is
we
asked
earlier.
How
do
we
actually
measure
what
provides
the
most
value
to
the
organization?
Like
what
does
that
mean
most
value?
Is
a
user
refused
its
viewings?
Is
it
ratings?
What
is
it
well
in
this
system,
the
more
SNT
that
has
been
staked
overall
by
all
the
developers
to
rank
their
d
apps,
the
less
SMT
there
is
in
circulation,
which
means
that
there's
a
higher
demand
per
SNC,
which
means
that
there
is
literally
and
I
do
mean
it
quite
literally
greater
value
for
SNC
holders
right.
A
So
you
can
be
sure
when
you're
looking
at
the
rankings,
okay,
Mehcad
our
ranks
firsts,
and
they
take
this
much
SNC.
That
much
SNC
has
effectively
been
removed
from
circulation
and
therefore
they
are
contributing
directly
to
value
that
I
also
am
a
part
of
as
an
SMT
holder.
So
it's
really
really
interesting.
You
know
that
was
the
state.
It's
really
interesting
that
this
is
a
sort
of
side
effects
of
how
we
design
decentralized
systems,
and
it
brings
me
back
to
what
I
was
saying
that
Beck
decentralization
is
the
goal
in
Gallet.
A
Arianism
is
a
great
success
metric
and
you
might
be
asking
you
know:
what
are
the
incentives
for
the
years
of
here?
Why
are
they
going
to
ever
interact
with
the
system
because
it
actually
costs
them
two
votes,
but
that's
kind
of
precisely
the
points.
But
if
you
think
about
our
current
systems
at
the
moment,
you
don't
want
to
have
to
actually
go
and
votes
on
every
single
Google
link
that
appears
in
your
search.
It
doesn't
make
sense,
and
nobody
really
wants
to
do
that.
A
But
it
costs
me
to
do
so,
and
that
seems
correct
and
that
cost
doesn't
go
to
a
centralized
company
like
Google
or
Amazon
or
Facebook.
It
doesn't
go
to
status.
It
goes
directly
back
to
the
developer
right,
so
it
becomes
in
a
sense
like
you
have
to
pay
to
complain
or
when
you're
up
voting.
Essentially,
it's
a
decentralized
patronage
system
for
the
apps,
which
is
really
really
interesting
again,
and
this
sort
of
brings
me
back
find
me
up
with
this-
is
the
wrong
place
to
the
difference
between
curation
and
discovery
right.
A
What
are
the
pros
and
cons,
yeah
and
obviously
like
I'm
deeply
biased
in
this
and
I?
Don't
know
huge
amounts
about
discovery
and
using
various
things
like
the
tribler
protocol,
which
is
a
way
of
indexing
content
on
BitTorrent
and
it's
already
sort
of
so
like
I,
say
I'm
a
little
bit
biased,
but
this
is
the
list
right
for
curation.
It's
easy
to
implement.
So
I'll
show
you
the
link
to
the
smart
contract
that
is
mostly
written.
There's
one
line.
A
That's
we
still
need
to
do
a
little
bit
of
work
on
it's
a
great
SNC
use
case,
and
that
also
applies
to
analytics
right.
So
chad
is
making
the
points
in
analytics
channel
earlier
that
in
facts.
You
know
when
we're
looking
at
ways
of
gathering
data
and
and
decentralized
fashion,
and
using
that's
usefully
in
terms
of
figuring
out
where
people
are
actually
getting
the
best
use
from
status.
Then
you
know
having
more
SNC
use.
Cases
is
really
really
useful
for
that,
because
it's
easy
to
measure
and
it's
all
publicly
available
on
the
chain.
A
Also,
the
the
UI
implementation
is
fairly
clear
right.
All
it
requires
is
a
new
tab
for
D
apps,
which
has
been
requested
by
a
lot
of
users
anyway,
because
they
say
that
they
tag
up
and
confuse
the
main
page.
Those
the
apps
just
need
to
have
a
small
little
like
amounts
next
to
their
name
and
then
two
little
buttons
put
down
votes
and
that
allows
the
community
to
both
interact
with
the
lists
and
make
it
clear
without
clogging
the
whole
thing
why
it
is
that
a
D
app
is
in
a
certain
position.
A
There's
one
smart
contract,
which
is
mostly
really
written,
and
you
know
one
of
the
difficulties
with
this-
is
that
it's
going
to
be
very
hard
to
personalize
and
filter.
But
you
know,
hopefully
with
one
smart
contract.
It
should
be
easy
to
either
extend
that
or
adopt
a
new
system
without
too
much
trouble.
A
A
Although
there
are
some
really
really
interesting
things
there
and
the
fact
that
we're
using
whisper
and
that
we
already
wrote
a
fairly
naive
version.
One
of
this
discovery
protocol.
There
has
been
a
lot
of
work
done
there
I'm,
just
not
too
aware
of
it,
so
I'm
very
open.
It's
a
discussion
about
what
the
actual
benefits
are
there,
because
this
is
a
bias
slide.
A
If
you
would
like
the
smart
contractors
here
me
and
Richard
have
been
working
on
it,
you
can
go
and
take
a
look.
It's
just
over
a
hundred
lines
of
code,
which
I
think
it's
pretty
exciting
and
the
conclusions
all
of
those
riders
that
tokens
are
just
these
abstract
data
structures,
guys,
which
means
that
we
can
perform
mathematical
operations
on
them
and
that's
nothing.
That's
particularly
me.
A
All
you
need
is
this
abstract
data
structure
that
you
can
perform
mathematical
operations
against,
which
is
those
curves
that
I
showed
you
and
then
get
the
results
back
and
use
that
in
a
decentralized
and
transparent
fashion.
So
the
aim
of
curating
information
without
any
central
authority
should
be
to
design
and
build
systems
that
cannot
be
cursed
by
any
group,
including
the
community,
especially
in
the
community,
and
such
a
vague
and
overused
term
anyway.
A
Tokenizing
information
is
a
great
way
to
do
that,
but
not
mean
you
take
it
literally,
as
that
complicates
both
the
incentives
in
the
UI
enormous
lis.
You
know
and
I
think
the
other.
The
other
sort
of
feature
here
is
not
only
what
are
the
incentives
for
the
user
because
they
actually
have
to
pay
to
votes
in
the
system.
What
are
the
incentives
for
the
developer?
A
Well,
developers
are
already
paying
for
things
like
Adwords
and
SEO
to
rank
highly
on
Google's,
so
like
it's
a
very
similar
system,
again
like
they
have
to
pay
to
rank,
except
that
money
doesn't
go
to
Google
or
a
centralized
entity.
It
gets
stored
in
the
smart
contracts
and
is
effectively
locked
out
of
circulation.
So
it
contributes
value
back
to
the
entire
community,
and
developers
stand
to
earn
a
significant
portion
of
their
back
through
community
votes,
whether
they're,
positive
or
negative.
So
it's
not
like
you
can
earn
money
back
from
SEO
or
AdWords.
A
If
lots
of
people
users,
you
might
be
able
to
do
that
with
Adwords
I.
Think
about
that,
but
certainly
not
SEO,
that
money
just
goes
to
Google
or
whoever's
doing
you
SEO
in
this
system.
You
actually
stand
to
earn
a
lot
of
what
you
stake
back.
If
you
can
get
the
community
to
actually
interact
with
us
importantly
and
I
know:
I've
mentioned
this,
but
I'll
just
emphasize
one
last
time
the
use
of
benefits
as
a
side
effects
of
the
optimal
and
decentralized
curation
of
information.
A
Much
like
they
do
now
right,
except
that
the
how
the
infos
are
ranked
is
Toki
transparent
and
they
can
affect
that
if
they
should
really
desire
to.
The
developer
also
stands
to
earn
back
that
significant
portion
of
what
they
stake
and
the
community
benefits
from
the
amounts
of
SNT
in
total
that
are
staked
and
the
destur,
because
that
is
removed
from
circle.
It's
a
great
token
sink.
Go
and
read:
movements
of
the
equity
tokens
like
that,
so
that
gets
a
great
piece
on
his
blog,
its
little
xca
and
that's
it.