►
From YouTube: Protocol Labs Research Overview - Evan Miyazono
Description
Originally recorded during the Berlin Developers Meetings from July 9-13, 2018.
A
A
So
when
you
see
that
this
is
where
you're
like,
that
is
what
it
represents,
the
idea
is
that
it's
a
blueprint
in
the
research
phase
of
things
and
what
I
plan
to
cover
today
is
enough
to
give
you
a
sense
of
why
protocol
labs
research
is
a
separate
group
or
like
a
unnamed
interface
within
protocol
labs
and
some
of
the
motivation
behind
what
protocol
labs
research
is
trying
to
do
and
to
be
fair.
This
is
all
my
interpretation
of
it,
but
I
think
that
it
gives
you
a
cohesive.
It
provides
a
cohesive
view
of.
A
The
answers
to
these
two
questions
by
explaining
why
we
often
they're
by
explaining
why
we
often
mix
described
protocol
labs
as
a
research,
development
and
deployment
lab
I'll
go
into
what
the
focus
of
research
is
as
well
as
answering
the
specific
question
towards
the
end
of
why.
Why
is
protocol
labs,
awarding
grants?
If
you
didn't
know,
protocol
labs
is
awarding
grants,
then
I'm
very
glad
you're
here
and
there'll
be
a
fair
number
of
details
about
that.
A
At
the
end,
as
like
a
meta
note,
this
90-minute
session
will
include
only
about
ten
minutes
of
me
talking
and
then
we
will
start
working
on
ideas
at
the
forefront
of
research
as
it
pertains
to
the
problems
that
we're
interested
in
and
by
we
I
mean
we
people
in
this
room,
not
we
protocol
labs
as
MP
us
or
royal.
We
so
jumping
into
answering
the
first,
a
that
you
may
have
seen
this
image.
A
A
A
There
is
such
a
thing
as
a
bad
idea,
and
not
all
ideas
should
be
turned
into
tools
that
people
have
access
to
either
because
they
would
be
very
bad
tools
or
they
would
violate
a
law
of
physics
and
you'd
spend
an
infinite
amount
of
time
trying
to
make
them
happen,
but
there
are
definitely
ideas
that
get
lost
along
the
way
and
to
highlight
two
of
them.
I
would
point
out
the
first
electric
car
from
1884,
and
that
was
a
technology
that
was
sad
to
sit
on
for
over
a
hundred
years.
A
But
even
worse,
potentially,
is
the
fact
that
the
first
steam-powered
motor
essentially
was
the
a
leo
pile
built
by
her
of
Alexander
in
the
1st
century.
Ad
and
the
first
actually
useful
steam
pump
came
of
millennium
and
a
half
later,
and
if
that
doesn't
make
you
feel
cheated
or
hit
you
in
the
gut
I
I.
Don't
know
what
would
imagining
us
being
that
far
along
that
much
further
along
technologically,
and
so
the
reason
that
we
built
we
are
building
this
research
development
and
deployment
lab.
A
A
Things
like
the
fact
that
every
human
should
have
access
to
all
of
humanity's
knowledge
and
have
whatever
it.
They
need
to
be
able
to
expand
it
on
that.
People
should
have
the
freedoms
in
the
digital
realm,
but
we
provide
everyone
in
their
physical
interactions
like
a
right
to
free
speech,
privacy
right
to
assemble
a
right
to
communicate
with
whoever
you
want
without
having
to
rely
on
a
third
party
and
also
the
belief
that,
if
you're
going
to
do
something,
you
should
do
it
well.
A
A
Decentralization
I
would
include
in
things
like
this.
The
ability
to
do
incentive
engineering
is
another
one
of
these
tools
and
even
though
we
don't
use
all
of
them,
it
doesn't
mean
that
we
believe
in
the
values
they
represent
any
less
and
from
those
we
build
infrastructure
that
incorporates
those
values
and
preserves
those
values,
and
so
from
an
overarching
perspective.
That
is
the
philosophy
for
research.
A
A
Where,
if
you
try
and
put
deadlines,
you
will
be
sorely
disappointed,
whereas
engineering
is
in
gold
ribbon
and
deadline-driven,
and
you
have
product
managers
and
you
have
people
who
understand
what
users
want,
and
so,
if
you
try
to
fit
many
of
the
things
that
protocol
abs
is
currently
doing,
you'll
see
that
much
of
it
looks
more
like
engineering
by
these
perceptions,
but
we
are
definitely
engaging
in
what
I
would
call
curiosity
driven
research
within
every
engineering
project,
as
well
as
beyond
an
external
to
the
engineering
projects.
I
would
not
consider
this
aimless
research.
A
Part
of
the
research
development
deployment
lab,
is
to
have
topic
driven
research
where
we
want
to
develop
toolkits,
improve
the
set
of
cryptographic
tools
to
enable
better
trustless
interactions
or
improve
on
the
the
incentive
systems
within
open-source
software
to
make
open-source
projects
more
sustainable,
and
so
this
could
be
small
projects
within
larger
project
or
small
questions
within
larger
projects,
or
it
could
be
these
nascent
independent
projects.
They
almost
seem
unrelated
until
you
take
a
step
back
and
reflect
on
the
overall
values
of
protocol
labs.
A
A
A
We
can
dive
into
these.
The
goal
was
to
source
some
more
from
all
of
you,
and
so
I'm.
Happy
to
the
plan
was
to
talk
through
the
development
of
one
that,
with
all
of
you
as
a
group
and
then
potentially
split
off
into
smaller
groups
and
have
as
many
groups
as
we
have
ideas
and
to
complement
the
open
problems.
These
are
in
the
protocol,
research,
repo
and
the
idea
is
to
have
a
place
to
discuss
these
ideas.
A
The
hope
is
that
eventually,
people
will
come
to
protocol
labs
and
come
to
this
repo,
in
particular,
to
post
questions.
They
have
have
discussions
here
and
make
use
of
the
compliment
of
the
open
problems,
which
is
the
request
for
proposals
program,
wherein
we
assign
some
amount
of
funding
to
individual,
open
problems
and
we'll
review
applications
and
actually
fund
people
to
work
on
promising
avenues
for
these
problems.
A
This
is
this:
has
the
potential
to
be
exceedingly
powerful
in
a
field
where
it
may
be
difficult
to
get
academic
funding
for
things
that
are
truly
impactful
in
the
end
to
engineering
projects,
whereas
we
know
the
problems
that
are
impactful
to
us.
We
have
funding
that.
We
can
put
behind
some
of
these
problems
and
optimistically
I
think
we
could
have
an
influence
over
the
directions
that
fields
move
and
take
them
from
solving
toy
problems,
to
solving
pivotal
problems
that
are
integral
to
some
future
engineering
efforts.
A
A
Call
it
a
draft
open
problem,
and
we
will
we
being
the
community
people
who
follow
this,
who
watch
this
repo
can
comment
on
the
issues,
refine
them
and
once
they
have
reached
the
level
of
formality
that
is
necessary
to
have
a
a
problem
statement
that
has
clearly
satisfiable
conditions.
Then
we'll
label
it
with
the
official
open
problem
tag
which
you'll
see
here.
We
I've
tagged
the
ones
that
are
open,
official,
open
problem
statements
as
well
as
ones
that
have
rfp's
open.
A
B
Have
a
problem
in
ipfs
right
now
that
dhts
scale
really
well
in
horizontally,
but
that's
no
matter
how
far
you
scale
horizontally.
It
doesn't
increase
the
vertical
scaling
of
dhts
this
more
specifically,
if
every
single
node
in
the
DHT
is
only
putting
like
a
constant
number
of
items
to
the
DHD,
it
skills.
B
Fine,
no
matter
how
many
notes
you
have
it
scales
well,
but
if
every
every
single
node
or
even
some
small
number
of
nodes
are
putting
an
unbounded
amount
of
content
to
the
DHT,
it
doesn't
scale
that
well,
which
is
fine,
because
ipfs
doesn't
put
a
lot
of
content
to
the
DHT.
But
what
this
does
put
a
lot
of
into
the
DHT
right
now
is
called
provider
records.
So
when
you
need
to
find
content,
you
have
to
look
it
up.
B
You
have
to
figure
out
who
has
this
piece
and
now
in
order
for
ipfs
in
order
to
provide
random
access,
we
provide
every
single
hash
of
a
graph.
So
if
you
have
a
graph,
that's
like
you
know
a
petabyte
of
files.
Each
of
those
is
one
256k
node,
so
like
millions
and
millions
and
millions,
and
so
if
you
want
to,
if
you
tell
IP
best,
to
provide
that
petabyte
of
data,
it's
gonna,
send
you
know
billions
of
DHT
puts
out
which
doesn't
work.
B
It
just
breaks
things,
so
we
need
to
find
a
solution
to
this
problem.
How
can
you
have
ipfs
users
that
have
petabytes
of
data
and
if
you
could
make
it
accessible
to
the
rest
of
the
network
without
completely
dossing
network?
This
thread
has
gotten
pretty
long.
It's
about
2
years
old.
Now
it
poses
the
problem
and
discusses
a
bunch
of
solutions
pretty
well,
but
definitely
still
an
open
problem.
C
B
Yeah
so,
like
I,
specify
the
problem
here,
I
suppose
I
go
through
a
bunch
of
potential
solutions,
talk
about
what's
good,
what's
bad
in
each
of
them
and
then
talk
about
it
a
little
bit
more
of
different
things
and
then
yeah.
The
problem
like
over
the
course
of
like
interacting
with
people.
The
problem
is
very
well
teased
out
here,
but
I
think
I
just
need
to
go
through
and
like
condense,
this
into
a
like
one
or
two
paragraph
precise
problem
statement
that
is
like
properly.