►
Description
This talk was given at IPFS Camp 2022 in Lisbon, Portugal.
A
Hi
everyone,
I'm
Nicholas
I'm,
founder
of
laptop
and
we're
going
to
be
hosting
the
decentralized
signs
and
Big
Data
Workshop
I'm
really
excited
for
it.
Okay,
so
I
think
there's
been
multiple
attempts
at
defining
decentralized
signs,
but
I
think
we
can.
We
can
derive
what
is
happening
and
where
this
word
is
coming
from
by
just
thinking
about
like
the
process
of
science
on
first
principles.
So,
first
science
is
a
knowledge
creation
process.
We
don't
know
something.
A
A
That's
where
the
concept
of
citations
and
papers
sort
of
comes
from
when
people
started
saying:
okay,
we
don't
need
to
respect
these
old,
like
antique
codexes
anymore.
We
let's,
let's
break
the
scientific
process
down
to
like
Atomic
units
and
start
exchanging
these
papers
and
cite
them,
and
then
we
have
the
open
science
movement,
which
was
I,
think
very
much
timed
with
the
invention
of
the
web,
where
at
CERN
there's
this
quote
from
Tim
Bernard
Lee,
where
he
says
well,
the
situation
was
just
so
dire
every
department
they
had
their
own
data.
A
It
was
really
hard
to
communicate
so
I
just
had
to
do
the
thing
and
combined
hypertext
and
other
formats
to
like
build
build.
What
we
now
know
as
the
web,
so
the
web
enabled
the
open
science
movement
and
now
with
web
3,
basically
distributed
storage
and
token-based
ownership
Concepts.
We
can
open
a
new
branch
of
the
science
movement
which
we
call
the
decentralized
science
movement
which
comes
with
some
additional
features.
A
So
some
of
the
new
paradigms
that
we
have
is
that
with
content
addressing
everything
is
citable
now
like
there's,
no
there's
no
such
thing,
as
is
this
something
that
I
can
cite
in
a
scientific
publication,
because
it
has
a
digital
object.
Identifier
or
not,
everything
can
be
cited
now
if
it's
content
addressable,
because
it's
there's
a
certain
guarantee
that
it's
static.
If
it's
pinned
consistently
and
then
we
have
a
another
concept
that
I'm
particularly
excited
about,
and
we're
going
to
be
diving
into
a
bit
today
with
the
bakalao
team,
which
is
content,
addressable
Transformations.
A
So
the
idea
that
if
your
content
is
out
there
in
the
web
and
is
directly
addressable,
you
can
think
about
the
process
of
doing
science,
especially
scientific
computation
as
a
directed
acyclic
graph,
where
you
can
just
where,
if
I
have
a
starting
point,
x
and
I
have
a
defined
transformation.
Some
piece
of
code
that
runs
in
a
standardized
environment,
I
can
deterministically
generate
an
output
y,
so
reproducibility
is
no
longer
opt-in.
Reproducibility
is
opt
out.
A
Where
tokens
enable
us
to
take
build
collectives
of
authors,
they
engage
in
a
scientific
project
and
then
the
citation
which
was
usually
the
metric
to
say,
okay,
who
did
most
of
the
work
who
did
like
who
contributed
less
so
it's
replaced
with
okay.
What
is
the?
What
is
the
contributor
token
balance
or
the
badge
number
that
one
scientist
has
in
that
particular
project
versus
the
next?
A
So
so,
why
does
this
matter?
Why
is
this
a
desirable
reform
of
the
way
that
science
works,
because
with
this
idea,
the
hope
is
that
we
can
give
everybody
the
opportunity
to
raise
funding
for
the
scientific
project,
irrespective
of
their
locations
and
irrespective
of
the
exact
nature
of
the
project,
whether
it's
totally
basic
funding
or
very
entrepreneurial?
A
If
you
can
post
your
project
description
online
and
you
can
give
a
multi-sig
address,
you
know
chances
are
you
have
a
higher
likelihood
of
finding
someone
who
really
cares
about
your
science
than
if
you're
staying
into
your
national
funding
infrastructure
systems,
then?
The
second
idea
is
that
we
can
enable
everyone
to
access
laboratory
services,
no
matter
where
they
are.
We
can
create
internet
protocols
that
enable
someone
to
book
a
laboratory
service
or
some
compute
service
and
pay
in
tokens,
irrespective
of
where
they're
based.
A
So
all
you
really
need
is,
is
a
good
idea
and
some
knowledge
about
what
needs
to
be
done
about
it,
and
then
third,
is
the
idea
that
we
can
share
materials
in
a
new
way
which
both
respects
the
inventor
they
generated.
That
idea,
but
also
makes
sure
that
that
that
new
knowledge
is
accessible
for
a
wider
audience.
A
So
at
laptop
we
started
out
thinking.
Okay,
with
this
change
in
the
communication
infrastructure,
that's
available
what
what
defines
an
effective
research
environment?
Could
there
be
a
completely
online
research
organization
because,
with
token
ownership
we
can,
we
can
think
about
like
these
type
of
questions
and
we're
not
the
only
ones
to
think
about
this.
So
this
is
a
slide
from
Juan's
talk.
A
couple
months
ago,
Ben
Reinhardt
wrote
about
a
private,
our
bus,
Reinventing
Discovery.
A
It's
a
really
really
good,
read
by
Michael
Nielsen
and
Adam
marblestone
Sam
Rodriguez
have
been
thinking
about
non-profit
startups,
for
science
funding
quite
some
time
and
there
are
multiple
historic
examples,
so
you
probably
know
about
Bell
Labs.
You
probably
know
about
DARPA,
but
I.
Think
one
example:
that's
a
bit.
Understudied
is
a
Cold
Spring
Harbor,
so
who
has
ever
heard
of
code
Spring
Harbor,
all
right,
one
yeah!
That's
that's
the
marine
biology
project,
of
course,
right
So,
So,
Cold,
Spring
Harbor
is
50
minutes
outside
of
New
York.
A
It's
a
commuter
rail
and
after
World
War,
II,
Cold,
Spring
Harbor.
You
you
hosted
these
into
this
highlight
into
disability
summer
camps
where
they
would
teach
scientists
about
phages.
So
bacteriophages
are
these
viruses
that
Target
bacteria
and
they're
pretty
harmless,
at
least
to
humans,
and
that's
that's
very,
not
very
beautiful,
because
it's
a
pretty
good
model
to
study
like
very
fundamental
biology.
So
at
that
time,
when
they
did
these
experiments,
they
didn't
even
know.
A
Think
something
we
can
learn
here
is
that
you
don't
need
a
lot
of
time
to
actually
come
up
with
really
new
ideas,
because
they
had
just
hung
out
here
over
the
summer
they
came
together,
they
formed
new
ideas
and
then
they
all
went
off
into
their
respective
environments
and
like
executed
on
them
until
they
could
re-gather
next
summer
and
what
if
the
internet
could
be
that
place
where
scientists
can
come
together?
They
hang
out
in
a
Lobby,
they
form
new
ideas
and
they
can
raise
funding
and
execute
on
them
using
using
new
tools.
A
So
an
effective
research
environment
is
a
place
where
you
have
access
to
Scientific
infrastructure,
where
it
needs
at
least
some
lab
space
or
some
tooling.
You
need
to
be
able
to
like
assemble
your
team
extremely
dynamically,
depending
on
the
need
of
the
idea
that
you
have
and
then
you
need
just
a
bit
of
funding
to
get
started
and
then,
once
you
see
that
you
know
your
idea
has
legs,
you
can
then
go
and
talk
to
the
existing
funding
agencies
to
to
get
some
real
support
and
the
Really.
A
The
reason
why
we
need
yet
another
science
organization,
I
believe
is
because
of
this
graphic
here,
which
shows
you
that
you
basically
have
migration
patterns
of
inventors
on
the
on
the
y-axis,
where
basically
above
zero
means
there's
a
net
influx
of
inventors,
where
an
inventor
is
defined.
As
someone
who
has
authored
a
patent
and
below
zero
means,
there's
a
net
out
e
flux
of
inventors
and
there's
really
just
one
country.
That's
a
net
importer
of
of
inventors
at
a
big
scale.
A
A
So
that's
how
we
arrived
at
lapdown
I
think
we're
not
alone.
There
are
a
lot
of
other
projects
in
this
space
and
I'm
really
excited
to
see
so
many
there.
This
is
a
this
is
a
slide
by
ultra
rare
Jocelyn's
team.
There
are
some
doubts
that
really
think
about.
Okay.
We
need
to
fund
signs
right,
so
scientists
can
apply
to
us
with
their
idea
and
then
we
have
places
where
science
gets
done.
A
A
Then
we
have
project
lion,
which
is
a
group
based
in
California
within
lapped
out,
together
with
Talent
out,
that's
looking
at
large
language
models
for
measuring
Discord
Community
Health,
and
then
we
have
the
knowledge
graph
lab
where
we
actually
are
going
to
see.
One
talk
later
from
from
Martin
from
lateral
about
the
work.
That's
been
happening
there
around
discourse,
graphs,
the
needs
that
we've
seen
with
projects
that
we
supported
and
and
helped
spawn
is.
A
We
can
bracket
them
into
two
buckets:
there's:
okay,
people
that
I
mentioned,
and
that's
really
big
right
now,
so
teams
need
support
to
safely
manage
funds
that
they
raise.
They
need
this.
Sometimes
they
have
legal
questions.
They
need
questions
with
raising
grants,
knowing
where
they
can
go
for
their
with
their
idea.
They
need
access
to
a
community
of
scientists
where
new
ideas
can
emerge.
That's
really
how
science
works,
it's
extremely
social
and
something
that
we're
looking
more
and
more
into
is
contributor
certificates.
A
So
if
you
raise
a
Bitcoin
Grant,
do
you
want
to
give
something
back
to
the
people
that
actually
helped
you
and
and
how
can
we
do
that
safely
and
then
on
the
other
bucket?
We
have
tools,
so
you
don't
only
need
talented
people
and
and
people
that
can
advise
you,
but
you
also
need
some
infrastructure
and
right
now,
when
you
look
at
the
basket
of
projects
that
we've
been
supporting,
those
are
mostly
happening
on
like
in
the
computational
realm.
A
So
computational
infrastructure
is
extremely
important,
but
we're
also
looking
into
laboratory
services,
and
this
is
just
one
example
for
an
application
that
we've
been
onboarding
to
to
to
do
some
signs,
and
especially
benchmarking.
This
is
equibind
It's,
a
graph
neural
network
that
allows
you
to
dock
small
molecules
against
proteins.
A
It's
basically
the
first
stop
that
ever
always
happens
when
you
try
to
design
a
new
drug
and
if
we
wrap
that-
and
if
you
stick
around
for
longer
in
this
Workshop
you're,
going
to
actually
be
able
to
use
that
because
we're
going
to
be
running
it
at
bakalao,
which
is
a
basically
a
Docker
inference
system.
That's
living
on
ipfs,
other
applications
that
we're
containerizing
and
making
accessible
to
scientists
are
rnac
tools,
histology
image,
analysis
tools,
antibody
language
models
and
so
on
so
mostly
biased
towards
biology.
A
A
The
way
that
we
plan
to
make
these
tools
available
at
scale
is
through
a
client
that
we're
developing,
where
baccala
is
a
core
dependency
where
scientists
all
over
the
world
can
install
the
client.
They
have
an
account
that
they
can
fill
top
up
with
a
balance
of
like
credits
where
the
credits
are
actually
tokens
and
that
it
can
spend
tokens
to
to
buy
laboratory
services
where
most
of
the
laboratory
services
will
be
computational
in
the
beginning.