►
From YouTube: Decentralization - with Pedro Domingos and George Hotz
Description
Fascinating discussion on how decentralization works and opportunities to create new paradigms with insights from 3 unique hackers, algorithm creators, and designers - April 23, 2022 on the Briar Patch Observatory podcast - https://briarpod.net/. Show notes here - https://wiki.opensourceecology.org/wiki/Briar_Patch_Observatory_Podcast
Marcin Jakubowski - https://www.ted.com/talks/marcin_jakubowski_open_sourced_blueprints_for_civilization
Pedro Domingos - https://homes.cs.washington.edu/~pedrod/
George Hotz - OpenPilot autonomous car software + https://en.wikipedia.org/wiki/George_Hotz
-----------------
What you see here at Open Source Ecology is an ambitious open source project for the common good. Take a minute to subscribe to our email newsletter (updates, workshops, etc): http://bit.ly/1LtcM44
A
B
C
D
F
A
A
B
A
A
Out
mouths
hanging
a
gap,
you
realize
what
you've
done,
dreading
you
weren't
too
loud,
but
your
fears
are
confirmed.
Your
alexa
chimes
according
to
the
terms
of
use
gendered
language
is
hate.
Speech
and
speech
is
violence
you'll
be
unable
to
purchase
anything
for
one
week
your
food
orders
have
been
cancelled.
A
A
D
D
E
B
D
A
D
F
A
H
Introduce
yourself
hey,
I'm
george,
I
come
from
old
school
hacker
culture,
you
know
hack,
the
planet,
information
should
be
free.
I
work
on
open
source,
self-driving
cars
and
it
turns
out
the
most
of
the
software
you
need
to
drive.
A
car
is
actually
also
the
software.
You
need
to
drive
a
person
so
probably
over
the
next
10
years,
we're
going
to
build
open
source
people.
A
A
D
B
F
D
D
E
D
D
B
A
F
Then
we
have
to
worry
about
whether
the
dynamics
is
going
in
our
direction
or
where
they
have
to
make
it
go
in
the
opposite
direction,
and
then
we
have
to
realize
that
there's
a
fight
that
will,
you
know,
always
have
to
be
fought
to
make
this
whole.
You
know
more
concrete,
you
know.
Let
me
just
give
a
very
simple
example,
which
I
think
is
very
topical
today,
which
is
you
know.
There's
an
army
work
better
if
it's
centralized
or
decentralized
right
and
a
very
centralized
army
does
not
work
well.
F
In
fact,
this
is
part
of
why
you
know,
supposedly,
ukraine
is
being
surprisingly
strong
against
russia,
and
supposedly
one
of
the
strengths
of
the
u.s
army
is
that
a
lot
of
decision-making
power
is
done
at
the
lieutenant
platoon
level,
as
a
result
of
which
you
know
it
can
be
nimbler,
et
cetera,
et
cetera,
right,
so
centralized
chains
of
command
tend
to
be
slow
and
have
a
disadvantage.
That
way.
On
the
other
hand,
a
very
decentralized
army
is
actually
not
an
army
and
does
not
win
a
war.
F
It's
just
a
bunch
of
independent
guerrilla
efforts
that
never
add
up
to
overwhelming
the
enemy
right.
So
what
what
level
to
have
in
different
respects
right,
so
you
want
the
centralized
decision
making,
but
at
the
same
time
you
want
what
this
lieutenant
learned
here
to
quickly
pass
to
that
lieutenant
over
there.
This
is
something
that
the
u.s
learned
in
iraq
right
was
to
have
these
keyboards,
where
you
know
platoons
would
share
what
they
had
learned
about
combating
ids
and
whatnot,
and.
F
F
And,
of
course,
the
army
is
that
one
extreme
of
the
spectrum,
but
you
know
just
any
company
right,
there's
a
there's,
a
there's,
a
reason:
they
have
a
hierarchical
organization
right.
A
hierarchy
is
really
things
that
are
centralized
in
degrees
right,
there's,
local
units
that
then
you
know,
coordinate
to
larger
units
and
how
we
organize
this
in
in
different
areas
is
actually,
I
think,
a
very
interesting
question
with
no
easy.
You
know
answer
let's
decentralize.
A
F
A
H
So
the
world
has
kind
of
trended
towards
centralization.
Maybe
modernism
is
the
peak
of
this.
We
have
our
grand
narratives,
I'm
a
post-modernist.
I
don't
believe
in
grand
narratives.
I
I
think
the
key
thing
is,
and
this
is
something
that
will
okay.
Humanity
used
to
live
in
tribes
of
200,
150,
200
people
and.
H
Yeah,
that
was
the
unit
that
could
support
life
now.
The
unit
that
could
support
life
is,
in
general,
probably
much
larger.
How
do
you
bring
that
number
down
to
one?
How
do
you
make
individual
individual
sovereignty
is
not
a
political
question,
but
a
technical
question.
People
claim
individual
sovereignty,
but
I
see
them
wearing
clothes
well,
where
they
get.
The
clothes
did.
H
H
A
H
H
H
If
you
could
figure
out
how
to
replace
the
state
with
a
better
version
and
not
better
because
of
ideology,
you
always
lose
if
you're
fighting
ideological
battles
but
a
better
technical
solution,
cheaper
power,
better
water,
you
know
more
available
stuff,
that's
actually
technically
better
than
what
the
state
is
capable
of
providing
the
state
loses
and
freedom
wins
and
I
think,
to
focus
on
it.
As
a
technical
question.
H
F
So,
but
just
just
to
interject
something
here,
please
there's
something
very
important
that
we
need
something
like
the
state
for
which
is
to
protect
ourselves
against
others.
Right
people
have
studied
things
like,
for
example,
the
tribes
in
the
amazon
and
and
they
have
an
uncanny
tendency
to
end
up
being
the
optimal
size
to
overwhelm
the
tribes
near
them,
so
they
keep
their
territory
while
not
having
so
you
know
you
don't
want
to
divide
your
share
of
the
wealth
too
much,
but
on
the
other
hand,
so
there's
there's
there's
the
pull
towards
individualism
right.
F
It's
like
you,
know,
you're
sovereign.
You
take
care
of
yourself,
you,
you
know,
you
have.
You
know
sole
access
to
whatever
territory
you
want,
but
then
you
have
to
associate
yourself
with
others,
so
it's
not
to
get
overwhelmed
by
the
ones
who
would
take
it
right
and-
and
you
know,
if
you
aggregate
yourself
with
more
people,
you
get
less
of
the
you
know
of
what
you
get.
D
D
F
H
And
I
think,
there's
a
very
fundamental
question
about
whether
offense
wins
or
whether
defense
wins
and
if
offense
wins.
I
agree
that
this
has
to
be
true,
but
when
you
look
at
cryptography
it
seems
like
defense
wins.
It
seems
like
it
is
much
harder
to
crack
codes
than
it
is
to
you
know,
make
you
know.
H
F
H
H
F
Where
the
office
defense
wins
actually
varies
by
area,
and
it
varies
over
time
right,
crypto
is
actually
a
great
example.
That's
at
first
time
right
now.
Yes,
defense
is
winning
yeah,
he
might
lose
again
and
and
and
it's
hard
to
predict-
and
certainly
in
the
past
one
or
the
other,
you
know
like
200
years
ago,
you
would
have
said
that
actually
offense
wins,
but
in
different
areas.
Right,
if
you
look
at
military
history,
you
go
through
period,
and
this
this.
F
F
Well,
we
are
not
in
the
scarcity
world
and
the
the
answer
to
that
is
that
you
know
human
beings
are
this
amazing
creature
that
has
an
unlimited
appetite
right,
so
we
could
have
a
million
times
more
wealth
than
we
did
today,
and
people
would
still
be
talking
about
redistribution
and
countries
would
still
probably
sure
invading
each
other
and
whatnot,
because
it's
human
nature
right.
It's
not,
you
know,
there's
always
a
limit
of
resources
and-
and
we
will
always
be
wanting
more-
you.
B
H
H
H
E
G
F
E
Very
much
competitive
and
all
that,
like
you,
can't
make
a
good
case.
I
don't
think
you
can
make
a
great
case
for
open
collaboration
being
the
way
of
the
world.
We
have
fighting
and
all
this
we're
talking
about
right
now,
so
we
need
to
evolve
as
humans,
so
can
technology
help
us
do
that?
Well,
the
first
prerequisite
is
meet
humans.
D
E
E
E
In
poland,
everything
is
like
all
gray,
like
real
material
scarcity,
I
had
to
wait
in
line
to
get
things
like
you
know,
for
an
hour
to
get
butter
or
or
meat
at
the
store.
Things
were
rationed,
so
you
got
to
solve
that
first,
once
we
can
solve
this,
we
can
free
up
people's
time.
Then
you
can
question
well
what
are
gonna
people
do
and
when
they
have
ample
spare
time,
but
we
have
to
start
at
the
basics.
When
then,
we
can
have
people
actually
start
pursuing,
what's
really
important
and
meaningful
to
them,
because.
E
A
F
A
Will
use
centralized
means
to
do
so
and
so
having
decentralized
technologies
or
anything
else
is
a
way
to
kind
of
have
some
type
of
autonomy
and
marcin.
You
grew
up
in
poland
under
communism,
page
where
you
grew
up
in
kind
of
bad
times
in
portugal.
Can
you
both
talk
a
little
bit
about
how
maybe
this
this
shapes
your
views
for
the
need
for
decentralization.
F
Sure
so
my
my
first
years
of
life
I
lived
in
their
fascist
dictatorship.
I
can't
say
that
you
know
I
was
too
small
to
really
understand
the
full
extent
of
it.
But
then
there
was
this
revolution
where
you
know
the
countries
here
towards
communism
and
then
back-
and
you
know
I
actually
learned
a
lot
during
that
period-
and
I
think
you
know
so
there's
this
central
dilemma,
which
is
you
need
the
state
to
protect.
You
from
you
know
the
people
who
would
kill
you,
but
then
the
state
will
kill
you
right.
F
And
this
is
the
problem
that
is
of
course,
in
my
view,
at
least
not
well
solved
by
fascism
or
communism
or
other
autocracies
right,
because
there's
a
conflict
of
interest,
so
so
the
rulers
are
maximizing
their
own
utility,
probably
at
the
expense
of
the
people
right,
so
you
need
assist.
So
at
the
same
time,
you
do
need
those.
You
know
some
amount
of
centralization
to
make
things
work.
F
You
know
whether
for
defense
or
or
justice
or
or
you
know
just
to
have
the
underpinning
of
markets
and
so
on
again,
there's
there's
a
lot
that
we
can
talk
about
there
and
and
technologists
change
it.
But
if
we
assume
for
us
for
a
moment
that
you
do
need
that,
then
you
also
have
to
answer
the
question:
how
do
you
prevent
the
people
who
are
then
in
the
position
of
being
the
rulers
or
the
whatever
you
call
them
from
from
exploiting?
F
You
know
their
position
and
oppressing
you,
which
is
what
one
way
or
another
ends
up
happening
in
every
autocracy
right,
and
I
think
that
you
know
the
good
old
u.s
constitution
is
actually
a
very
nice
example
of.
How
do
you
do
this
kind
of
mechanism
design
right?
I
think
the
problem
with
a
lot
of
idealists,
particularly
on
their
left,
is
that
they
want
to
have
a
system
of
government
that
is
based
on
the
rosiest
possible
assumptions
about
human
beings.
F
If
we
were
all
just
you
know,
peace
loving,
you
know
brotherhood
kumbaya,
maybe
socialism
would
work,
but
we
aren't
the
thing
that
the
founding
fathers
obsessed
about
and
very
properly
so
and
and
with
great
success
and
with
the
beneficiaries
is
we
need
to
have
a
system
that
will
work
even
under
the
most
pessimistic
possible
assumptions
about
human
nature
right.
What
if
we
have
people
who,
because
you
will
have
someone
in
power
who
wants
to
take
too
much
power?
Who
wants
to
take
advantage?
You
know,
etc,
etc.
F
Right-
and
this
is
where
things
like
you
know-
separation
of
powers
and
so
on,
and
so
on
come
from
so
that
you
know,
if
we,
you
know,
I'm
a
computer
scientist
like
so
thinking
in
terms
of
mechanism
design,
I
think,
comes
very
naturally
right.
You
can
think
of
the
constitution
as
being
an
algorithm
for
how
to
run
things,
and
it
was
amazingly
insightful
and
the
the
part
of
a
lot
of
hard
work
for
the
time.
F
I
think
you
know,
of
course
you
know
it
does
have
some
bugs,
but
the
the
interesting,
I
would
even
say,
fascinating
question
for
us
today
is
that
with
the
technology
that
we
have
today,
can
we
actually
design
something
better?
In
particular,
I
think
of
interest.
You
know
to
this
conversation,
something
that
actually
requires
less
centralization
where,
while
potentially
delivering
better
results-
and
I
think
the
answer
to
that
is
probably
yes,
you
know,
I
think,
there's
a
lot
of
experiments
at
various
levels,
from
cities
to
parties
to
whatnot.
F
You
know
I
mean
lots
of
other
things
going
on
around
the
world.
I
don't
think
anything
has
really
caught
fire
yet,
but
I
think
sooner
or
later
we'll
it
will
and
also
you
know,
just
you
know,
here's
one
thought
to
bear
in
mind.
I
think
you
know,
as
a
young
researcher,
I
see
with
a
lot
of
worry
how
autocracies,
like
china
in
particular,
are
you
know,
running
full
steam
ahead
with
using
ai?
F
H
All
these
things,
though,
are
political
solutions
like.
I
don't
think
that
you'll
ever
win
politically.
Here's
here's
my
basic
theory-
and
it's
like:
why
do
people
want
to
oppress
others?
Why
do
people?
Where
does
that
desire
come
from,
and
I
think
it
basically
comes
from
a
desire
for
slaves?
I
think
that
90
of
people
have
honestly
asked
would
love
to
have
slaves
right,
they'd
want
to
be
a
slave.
They
would
love
to
have
sex
right.
H
A
H
F
I
I
actually
agree
with
that
that,
to
a
large
degree
in
the
following
sense,
I
think
if
you
do
the
following
thought:
experiment,
which
I
think
is
a
good
starting
point,
is
last:
ask
everybody
if
they
would
like
to
be
king
of
the
world
right,
not
slaves,
but
just
the
world
exists
to
serve
you
right.
Most
people,
I'm
guessing,
would
probably
deep
down,
or
at
least
or
even
very
openly
say,
yeah
sure
right.
Why
not
right?
F
Why
not
have
the
world
serve
me
right,
but
then,
but
then
of
course,
and
so
then
the
slaves
are
just
the
means
of
that
then
you're
right.
If
you
have
robots,
then
you
don't
need
the
slaves,
but
also
you
need
to
protect
yourself
from
others,
and
also,
I
think
there
you
know,
there's
a
social
power.
You
know
because
of
this
right
there's
you
know
like
in
order
to
get
resources,
you
need
to
dominate
over
others,
and
that
sometimes
ends
up.
F
You
know
becoming
it's
its
own
goal,
but
bottom
line
right
oppressing
people
is
actually
very
costly,
exactly
get
what
you
want
without
oppressing
them.
That's
much
preferable.
In
fact,
one
of
the
chief-
maybe
even
the
chief
there's
many,
but
I
think
the
chief
you
know
the
you
know,
handicap
of
of
authoritarianism-
is
the
cost
of
oppressing
people
which
democracy
does
not
have,
or
at
least
that's
to
a
much
lower
degree.
H
C
H
E
D
G
E
B
H
I
buy
into
the
basic
neo-reactionary
tenet
that
says
we
have
become
so
much
worse
at
political
algorithms
in
the
last
300
years,
like
300
years
ago,
people
actually
thought
about
this
stuff
because
it
was
life
or
death.
Now
we
live
in
a
technological
paradise
and
you
know
we
all
push
through
stupid
ideas
and
like
it
doesn't
matter,
you
know.
A
But
these
are
of
course
antithetical,
although
most
people
just
say
them
in
the
same
breath,
and
the
constitution
was
designed
not
exactly
for
separation
of
powers,
but
that
there's
sharing
of
powers
such
that
no
branch
is
able
to
conclusively
determine
the
extent
of
their
own
powers
right.
So
you
can't
have
just-
or
at
least
it
was
designed,
so
you
couldn't
have
just
one
branch,
just
kind
of
dominate
all
the
others
and
I'm
curious
when
it
comes
to
you
know
technology
and
decentralization.
A
F
Then
there's
this,
you
know
positive
feedback,
loop
right
and-
and
you
could
say
that
in
many
ways
that
positive
back
loop
is
a
good
thing,
because
we
are
now
better
served
than
we
were
before
right,
but
also
introduces
certain
dangers
that
you
have
to
guard
against.
At
the
same
time,
right
I
mean
these
days
almost
across
the
spectrum.
You
have
these
people
saying
that
the
big
tech
companies
are
evil
because
of
their
centralized
power
and
whatnot.
But
but
the
thing
to
realize
is
that
this
centralization
has
actually
gone
hand
in
hand.
F
We
decentralization
at
another
level.
So,
for
example,
thanks
to
google
and
amazon,
more
small,
you
know
authors
merchants,
you
know,
makers
of
things
can
now
make
a
living
than
ever
could
before.
So
what
we
have
was
like
you
know.
We
have
this
simultaneous
centralization.
Where
there
are
these
behemoths,
like
google
and
amazon,
that
actually
support
an
enormous
decentralization
and
that,
like
anybody,
can
sell
things
any
anybody
can
put
out.
You
know
a
podcast,
for
example,
that
will
have
you
know
an
audience
of
people.
You
know
potentially
arbitrary
people
around
the
world
that
couldn't
before.
A
When
you
said
that
it
reminded
me,
basically,
it
sounded
almost
like.
You
said
that
the
all
these
platforms
have
almost
created
a
new
commons
for
people,
and
I
think,
10
years
ago,
a
lot
of
people
saw
that.
Oh
you
know
it's
great.
That
google
has
put
all
the
all
these
things
out
for
anyone
to
use
and
all
these
platforms
that
appeared
to
be
neutral
kind
of
like
a
commons,
was
back
in
the
day,
but
now
they're
kind
of
exerting
themselves
over
these
commons
and
people
don't
like
that.
F
F
It
would
just
be
different
platforms,
but
that
creates
ideally,
this,
you
know
even
playing
ground
for
very
large
scale,
decentralization
to
happen,
but
then,
of
course,
a
lot
of
interesting
things
happen
at
the
margins
is
where
not
even
if
they
start
at
the
margins
of
so
the
platform
is
centralized
like
amazon
or
google,
the
users,
like
the
merchants
and
the
you
know,
the
buyers
are
very
decentralized
like
never
before,
but
then
the
question
becomes.
Where
does
the
platform
end
and
the
user
begin
or,
like
the
you
know,
the
you
know
think.
C
F
Even
an
operating
system,
right,
like
is
the
browser
part
of
the
operating
system
right
or
like
the
platform,
is
content.
Moderation,
part
of
the
platform
once
content
mod
and
your
content
moderation
didn't
need
to
be
part
of
the
platform
when
the
platform
was
small.
You
know
think
of
all
this
social
media
right
at
some
point.
The
content
moderation
becomes
absolutely
essential
because
there's
too
much
information
for
it
to
work
without
content
moderation,
but
at
that
point
it's
no
longer
just
the
platform.
It
actually
has
started
to
be
a
centralized
thing
right.
H
The
notion
that
these
platforms
are
like
powerful
or
monopolies
just
never
seemed
true
to
me.
If
you
think
about
the
railroad
or
the
electrical
grid,
these
are
centralized
monopolies
that
would
cost
huge
amounts
of
resources
to
reproduce
the
minute.
Someone
makes
a
search
engine
better
than
google
they'll
take
over
the
search
engine
market.
In
five
years
right,
I
wrote
a
blog
post
about
vampire
attacking
twitter,
I
think,
with
a
billion
dollars
you
could
take
over
twitter,
you
don't
need
to
buy
twitter.
H
A
H
Oh,
so
you
can
overcome
network
effects
by
incentivization
or
like
liquidity
attacks,
so
like
the
famous
one
in
crypto.
Is
that,
like
sushi
swap,
took
all
the
liquidity
from
unispot
by
basically
incentivizing
the
lps,
but
how
hard
would
it
really
be
to
incentivize?
You
know
the
10
000
top
people
on
twitter
to
move
somewhere
else.
F
Very
hard,
actually,
so
I
I
you
know,
I
I
I
actually
agree
with
you
that,
and
this
is,
I
think,
not
widely
appreciated
that
it's
quite
possible
to
create
a
search
engine
that
beats
google
or
you
know,
a
social
network
that
beats
student
and
whatnot.
But
I
think
you
are
greatly
underestimating
the
difficulty.
F
Can
make
another
google,
it's
just
a
search
box?
It's
easy
to
switch!
It's
not
like
windows,
where
you
know,
you'd
have
to
throw
out
all
your
software,
but
now
we
know
you
know
as
far
as
mentioning
there's
the
network
effect
of
data
right.
You
know
you
you,
you
know,
you
know,
google
has
you
know
10
orders
matter
to
more
data
than
you,
so
your
ranking
learning
algorithm
has
to
be
10
hours
of
magnitude
better
for
you
to
compete
with
it
and,
and
that
is
not
available
right.
F
So
so
you
know
like
if
you
I
talk
to
folks
at
bing
who
say-
and
actually
you
know
agree
with
them
from
what
I've
seen
that
technically
being
is
actually
comparable
to
google
or
better,
but
that
they
they
have
no
hope
they've
given
up
on
ever
trying
to.
You
know
dethrone
google,
because
the
advantage
that
google
gets
from
the
extra
day
that
it
has
and
therefore
producing
better
search
results,
is
just
too
big
same
thing
with
twitter.
F
H
Again
from
scratch
somewhere
else
and,
of
course
you're
not-
and
this
is
why,
in
the
twitter
case,
what
you'd
have
to
do
is
you'd
have
to
copy
the
graph
right,
like
you'd,
scrape
twitter's
entire
graph
you'd
put
twitter's
graph
somewhere
else,
it's
basically
identical
to
use,
but
you
get
incentivized
to
tweet.
I
also
think
that,
of
course,
bing
is.
H
H
I
don't
think
twitter
could
could
stop
that,
because
it's
indistinguishable
from
using
the
service
bing
will
never
beat
google
because
being
it's
even
but
it's
not
google
like
they
built
the
same
thing.
They
built
a
box
you're,
never
going
to
build
a
better
box
than
google.
But
could
you
build
a
nlp?
F
Well,
that's
that's
funny.
You
should
say
that,
because
that
was
actually
the
big
push
at
bing
in
the
early
days
you
know
of
steve
ballmer.
I
mean
I
knew
the
people
working
on
this.
It
was
exactly
we
are
going
to
beat
google
by
being
an
nlp
driven.
You
know,
dialogue.
Ask
me
questions
engine
that
was
their
whole
idea
of
how
they
were
going
to
beat
google
and
they
didn't.
F
No,
I
mean
the
train
test
is
meaningless.
It's
already
been
passed
in
in
many
meaningless
ways,
and
but
I
mean
that
without
going
there,
I
think
the
question
of
when
will
the
technology
be?
There
is
a
good
one,
but
the
thing
to
realize
is
that
no
one
understands
that
better
or
is
pushing
more
on
that
technology
or
is
farther
advanced
in
it
than
google
itself.
H
H
F
The
whole
idea
of
page
rank
that
made
google
successful
is
that
the
best
source
for
whether
a
page
is
good
is
the
people
who
decide
to
point
to
it
or
not,
and
then
the
people
who
point
to
those
and
whatnot
so
google
is
powerful,
because
it's
a
decentralized
search
engine,
the
the
computation
happens
in
a
centralized
fashion
but
doesn't
need
to
write,
but
that's
almost
like
a
secondary
technical
issue.
The
point
is
that
google
is
powerful
because
it
has
everybody's
brain
working
on
ranking
the
the
the
pages
as
opposed
to
tfidf,
where
you
know
you're.
F
Oh,
I
think
google
will
inevitably
go
down
at
some
point.
I
think
all
these
you
know
I
I
I
have.
I
cast
a
skeptical
eye
on
all
these.
Like
antitrusts,
you
know
lawsuits
against
tech,
companies
and
threats
of
them
because
they
always
end
the
same
way,
which
is
you
know
the
company
loses
its
dominance
before
the
yeti.
Trust
lawsuit
runs
its
course
and
wastes
millions
of
dollars
and
all
of
these
tech
companies.
This,
I
think,
is
an
interesting.
You
know
a
question
for
us
and
and
many
others
is.
F
F
One
is
that,
if
I
come
up
you
know,
google
can
only
you
know
they
try
to
quarter
the
market
in
machine
learning,
researchers
and
that
they
almost
did
at
one
point
now
now
it's
impossible,
and
you
know,
sooner
or
later
someone
is
going
to
come
up
with
a
machine
learning
algorithm
that
is
100
times
better
than
theirs,
and
at
that
point
I
only
need
a
hundredth
of
the
data
to
be
competitive
right
and
they
understand
this
right.
They
are
terrified
of
this,
but
sooner
or
later
it
will
happen.
This
is
one
aspect.
F
Another
aspect
is
that
they
make
their
money
by
targeting
ads
right
and-
and
there
are
you
know,
ad
block
is
like
there
are
many
interesting.
You
know
arms
races
going
on
in
the
world
right
now
in
technology
in
particular,
and
one
of
them
is
but
but
this
one
actually
surprisingly,
in
its
early
stages,
when
it
shouldn't
be,
is
between
ad
blockers
and
ad
sellers
right.
But
you
know
we
were
talking
about
offensive
defense.
F
H
Ads
is
another
form
of
where
you
are
deviating
from
the
ideal
function
right.
What
you're
the
ideal
function
of
what
the
person
actually
wants?
Well,
you're,
not
just
taking
that
into
account
you're,
also
taking
account
what
dyson
vacuums
wants
when
they
bought
this
app
right.
So
so
they're
already
like
they're
already,
you
know
sub-optimal
right.
Could
I
I
could
build
the
same
thing
as
google
put
no
ads
on
it,
and
people
probably
use
it.
No.
F
F
H
F
F
And
what
this
means
is
that
again,
this
makes
google
vulnerable
to
someone
who
basically
says
no,
I'm
just
going
to
have
fewer
ads
again
right.
They
have
this
reward
function
that
causes
them
to
to
step
by
step
with
great
resistance.
Put
more
and
more
adds
at
some
point
again.
I
think
it's
hard
for
them
to
fight
that,
but
at
some
point
creative
destruction
comes
into
play
and
you're
right.
They
get
bad
enough
that
somebody
else
you
know
who
has
a
good
enough
algorithm,
maybe
better
whatnot,
but
also
gives
you
less
ads
and
less.
F
F
Right
and
like
they,
what
was
the
name
of
this
company
that
they
bought?
You
know,
I
don't
know
like
that
that
that
created
this
idea
of
you
know
of
having
you
know
ads
placed
right,
but
what
they
bought
them
and
that
became
the
foundation
of
their
business,
but
they
made
a
huge
change,
which
was
they
used
to
intersperse
the
sponsor
than
their
real
results,
without
telling
you,
which
was
which,
as
a
result,
you
didn't
because
they
thought
they.
C
F
A
A
To
ask
the
question
about:
will
the
financial
incentives
for
decentralized
technology
and
I'm
going
to
ask
marcin
first
because
he's
not
in
the
software
business
he's
more
in
agriculture,
heavy
industry?
It
might
be
a
little
bit
different
and
I'm
curious
if
the
financial
pressures,
if
they're
different-
and
it
seems
to
me
and
correct
me
if
I'm
wrong,
marcia
but
you're
kind
of
doing
this
kind
of
out
of
principle.
You
probably
could
be
doing
something
that
gives
you
a
greater
monetary
return,
but
you
seem
to
have
a
mission
to
deliver.
B
E
A
B
D
D
E
E
E
B
E
E
D
E
D
D
E
E
F
B
E
H
B
D
D
D
B
E
D
D
B
B
F
I
think
there's
a
very
important
factor
here,
which
is
customization
right.
Yeah
john
deere
can
probably
make
a
million
copies
of
the
same
tracker
more
efficiently
than
you
know,
a
million
you
know
farmers
making
each
their
own
tractor,
but
once
you
want
a
different
customized
tractor
for
each
farmer,
it's
you
know
at
least
potentially
the
individual
farmers
have
the
advantage
provided
that
they
have
the
tools
to
make
things,
because,
among
other
things,
they
know
what
they
want
better
right.
Why
do
you
wait?
Why
do
you
want
customization?
F
No,
I,
when
you
don't
always
want
customization,
but
you
very
often
do,
and
when
you
do
is
what
I'm
saying
when
you
want
customization,
then
decentralized
potentially
has
a
big
advantage,
because
you
know-
and
but
you
know
we
can
do
you
know
to
go
back
to
what
martian
was
saying
earlier.
We
can
do
this
sort
of
experiment
of
saying,
like
you
know,
suppose
you
have
a
world.
You
know
to
oversimplify
a
little
in
which.
F
H
F
Better
like
well,
I
mean
it's.
My
point
is
that
it
can
actually
go
both
ways
right
because
there
are
things
favoring,
one
or
the
other
right,
so
it
does
make
a
lot
of
the
so
amazon's
one
of
their
core
strengths
is
the
huge
you
know,
logistics.
You
know,
network
and
expertise
that
they
have,
and
that
would
disappear.
This
would
make
that
disappear
right.
So
that
puts
amazon
in
danger.
F
On
the
other
hand,
a
lot
of
the
expertise
that
they
have
more
on
the
software
level
might
even
you
know
their
advantage
compared
to,
for
example,
walmart,
would
be
even
bigger
in
such
a
world.
You
know
they
would
be
the
company
that
provides
all
the
software
and
maybe
even
the
3d
printers
and,
like
you
know,
the
organization
and
so
on,
to
to
let
people
do
this.
F
You
know
like
often
by
scrambling
like
microsoft.
You
know
they
wrote
the
wave
of
pcs
and
they
just
barely
managed
to
adapt
to
the
internet
and
facebook.
You
know
managed
to
just
barely
adapt
to
mobile
right.
Facebook
could
have
died
when
mobile
came
along,
they
almost
did,
but
they
didn't
right
and
now
they
dominate
right,
but
they
will,
sooner
or
later
be
a
wave
that
they
fail
to
adapt
to
right,
and
I
think
there
will,
sooner
or
later
be
a
wave
that
amazon
fails
to
adapt
to.
F
It
could
be
this
way
where
things
you
know
just
you
know,
become
more
and
more
made
in
the
home
or
not
right
and
again,
there's
people
at
amazon
who
you
know
whose
job
is
to
think
about
things
like
this
and
try
to
be
ahead
of
them.
What
often
happens-
and
again
this
is
the
downside
of
centralization-
is
that
there
then
there
are
people,
you
know
it's
that
saying.
F
I
don't
know
if
you've
heard
this,
that,
like
any
company
large
enough
to
have
a
research
lab,
is
too
large
to
listen
to
it
right,
so
so
they
could
fail.
Even
though
you
know
there
are
people
who
inside
them,
who
see
what
should
be
done,
but
the
at
that
point,
the
forces
you
know
for
the
status
quo
inside
the
company
are
too
large
and
of
course
there
are
countless
examples
of
big
fails
like
this.
H
Regardless
of
like
which
companies
specifically
see
or
fail,
I
see
software
and
hardware
as
being
very
different,
because
software
has
zero
marginal
cost.
The
cost
to
replicate
software
is
zero.
The
cost
replicated
tractor
will
always
be
something
and
to
use
3d
printers.
As
an
example
sure
we
went
from
the
ten
thousand
dollar
stratasys
era
to
the
reprap
open
source
era
to
the
the
best
3d
printer
is
now
a
thousand
dollars
from
china
error.
It's
not
open
source.
D
E
E
D
D
Here
we're
making
silicon
and
microchips,
so
that's
it's
not
too
far
out
I
mean
technology
is
getting
better
and
better.
It's
miniaturizing,
so
we're
kind
of
thinking.
Okay.
What's
can
you
actually
show
that
yeah
you
can
create
advanced
civilization
on
very
small
scales?
That's
that's
my
take
on
this.
We
haven't
seen.
F
H
F
Are
you
making
microchips
and
they,
but
you
know
just
to
finish
that
thought
there
are
a
few
things
more
centralized
now
than
microchip
production
yeah
right,
because
the
cost
of
the
fabs
is
so
high
right.
It's
like
tens
of
billions
of
dollars
so
over
time,
what
you
have
is
fewer
and
fewer
of
them.
Yeah.
B
B
E
F
Well,
we've
touched
on
a
bunch
of
them
so
far
right
so
and
then
there's
more,
but
again
you
know
without
going
back
over
those.
I
think
one
thing
that
that
I
think
marcin
briefly
touched
on,
but
I
think
is
very
important.
Is
this
question
of
competition
versus
cooperation
right?
It's
and
again
what
you
actually
have
most
of
the
time
in
most
domains
is
not
pure
competition
and
not
pure
cooperation
is
actually
have
what
sometimes
is
called
coopetation
right.
It's
a
mixture
of
competition
and
cooperation,
even
among
competitors.
F
You
know
companies
that,
like
you
know,
the
story
is
like
oh
they're,
at
each
other's
throats,
less
visibly
they're,
often
cooperating
because
they
actually,
you
know
so
think
of
you
and
me
writing.
You
know
as
two
companies
or
two
people
right.
There
are
areas
where
we
have
common
interests
and
therefore
it
makes
sense
for
us
to
cooperate
in
them
and
there
are
areas
where
we
have
opposite
interests
and
therefore
we
must
compete,
and
now
that
you
know
I
think
the
larger
pictures
are
like
you
will
get.
F
You
know
more
centralization
or
more
decentralization
depending
on
on
how
much
how
powerful
each
of
these
is,
and
what
often
happens
is
that
you
know
we
is
that
again,
people
always
see
competition,
club
cooperation
as
opposites
as
competing
forces,
but
ironically,
in
many
ways
it's
the
opposite.
It's
that
competition
is
the
mother
of
cooperation.
F
The
only
question
is
whether
it's
with
an
external
invader
or
with
each
other
when
there's
an
external
invader,
they
get.
You
know
they
are
very
good
at
actually
fighting
wars
right.
This
is
actually
the
most
potent
thing
about
what
happened
with
the
us.
There
is
that
they
learned
this.
You
know
they.
They
managed
to
bring
learn
helplessness
to
the
to
the
afghanistan
war
fighters
right.
F
When
there
is
an
external
threat,
like
you
know,
britain
or
russia
or
america
or
whoever
the
next
one
might
be,
then
it
makes
sense
for
them
to
cooperate
so
that
they
have
critical
mass
to.
You
know
to
face
down
that
external
threat,
and
once
that
external
threat
is
just
down,
they
go
back
to
fighting
with
themselves.
F
So,
and
I
I
think
this
is
an
example,
so
that
goes
on
in
a
lot
of
different
arenas
right
and
then,
as
the
cost
of
different
things
changes.
You
know
the
competition,
you
know
changes
more
towards.
You
know
larger
cooperation
smaller,
but
but
that's.
I
think
that
the
axis
along
which
you
have
to
look
at
this.
F
Seeing
more
cooperation,
I
mean
this
is
the
magic
of
the
internet.
Right
is
that
it
has
made
cooperation
on
a
scale
not
just
in
terms
of
numbers
of
people,
but
like
my
ability
to
find
someone
to
cooperate
with
me
across
the
world.
Right
is
something
on
a
different
scale
than
it
was
before
right,
and
so
this
really
does
create.
You
know
an
opportunity
for
an
exponentially
larger
number
of
connections
to
make
right.
F
Think
of
this
in
terms
of
game
theory
right
and
of
like
finding
non-zero
games
to
play
right,
the
the
the
internet
makes
it
possible
to
find
way
more
non-zero
games
to
play
than
before
right.
Some
of
those
non-zero
games
will
be
like
small
trades.
You
know
like
I
make
something
on
etsy
and
you
buy
right.
F
So
so
the
internet
did
create
this
enormous
opportunity
for
more
collaboration,
some
of
which
will
be
small
and
some
of
which
will
be
on
a
very
large
scale,
and
you
know,
and
and
they
may
change
right.
So
you
know
to
go
back
to
the
amazon
example.
One
of
the
main
threats
to
amazon
today
is
shopify.
Right.
Shopify
actually
is
a
very
interesting
threat
to
amazon
from
this
point
of
view,
because
they
are
less.
What
shopify
does
is
that
they
just
provide
the
service
to
you.
F
Yeah
and
they're,
you
know
they're,
like
you
know,
in
the
tens
of
you,
know
billions
of
dollars
company
at
this
point
and
and
growing
rapidly.
What
shopify
does
is
it
it
provides
to
you
if
you
are
a
merchant
of
anything
and
you
want
to
go
online,
it
provides
to
you
all
the
tools
and
all
the
infrastructure.
You
sell
your
goods
online
under
your
own
brand
name
as
opposed
to
amazon.
That
says
you
will
sell
these
things
on
amazon
right.
F
So,
in
a
way
it
provides
to
you,
you
know
all
the
goodies
that
amazon
does
without
the
downside
of
like
oh
no
amazon
is
taking
control
and
reminding
my
data
to
compete
with
me
and
whatnot.
All
of
those
worries
that
merchants
you
know
justifiably
have
about
amazon.
You
know
shopify
makes
them
go
away,
which
is
precisely
why
it's
grown
so
much
right.
So
shopify
is
a
more
decentralized
version
of
amazon,
which
maybe
now
makes
more
sense
for
the
for
the
merchants
right
and
therefore
is
greg.
A
Well,
one
of
my
one
of
my
big
questions
is
basically
about
vc
funding,
and
you
know,
if
I'm
a
venture
capitalist.
I
want
to
invest
in
a
company
that
can
grow
to
be
massive,
that
has
a
nice
moat
around
it,
and
so
I
would
think
that
there's
not
a
whole
bunch
of
venture
capitalists,
perhaps
beating
down
the
door
of
marcin
trying
to
give
him
funding.
F
I
don't
know
if
you've
heard
this
concept
of
the
kill
zone
around
the
tech
giants
right.
The
kill
zone
around
the
tech
giants
is
supposedly
this
zone
of
things
that
are
that
vcs
won't
invest
in
because
the
threat
from
the
tech
giant
is
too
great
right,
and
if
you
do,
this
amazon
will
come
and
squash
you
or
apple
or
google
or
whatever.
F
Therefore,
we're
not
going
to
invest
in
it,
and
then
you
can
and
which
is
a
shame
right,
because
then
then
there's
less
innovation
there
and
then
the
people
lose
out
right
and
and
then
outside
of
the
kill
zone.
Things
that
look
distant
enough
from
what
the
tech
companies
are
doing,
then
those
maybe
can
grow
more
and
have
funding,
and
maybe
one
day
one
of
them
will
go
to
be
a
threat.
One
of
the
you
know
those
tech
companies,
but
you
know
inside
the
kill
zone.
F
Life
is
very
hard
right
and
and
the
thing
about
this
concept
of
the
kill
zone
is
that
it's
a
very
popular
one,
but
there
are
so
many
exceptions
right
again.
Shopify
is
a
good.
You
know
this
shopify
started
as
a
very
small
startup
right,
so
you
can
actually
be
right
inside
the
tech,
giant's
kill
zone
and
yet
still
be
able
to
grow
and
defeat
it
right
and
part
of
this
is
the
moat
right.
You
know,
you
know
you
can
have
a
moat.
F
So
you
know
at
the
end
of
the
day,
I
think
you
know
if
you're
the
vc
right
you,
the
reason
you
need,
I
mean
having
modes
is
always
good,
but
but
the
reason
why
emote
is
essential
is
that
you
are
going
to
sink
a
lot
of
money
into
this
company.
Like
say,
uber
right,
tens
of
billions
of
dollars
have
been
sunk
into
uber
in
the
short
term.
F
This
is
only
worth
it
if
you're
going
to
make
a
lot
of
money
over
10
or
20
or
30
years.
Otherwise,
it's
not
work,
otherwise
it
won't
pay
off
right
because
there's
this
you
know
discount,
you
know
factor
you
know
for
the
future.
So
if
you
don't
see
a
potential
moat,
there,
you're,
probably
not
gonna,
invest
in
that
to
start
with
right
or
or
a
big
network
effect
or
or
some
combination
of
the
two
right,
and
you
know,
for
example,
you
know
to
to
take
the
example
of
uber.
F
You
know
made
a
ton
of
money
for
its
vcs,
despite
the
huge
amount
of
investment
that
went
there
so
so
most
are
important,
but
they,
I
think,
are
are
increasingly
hard
to
find
which,
if
you
believe
in
decentralization
is
good
right,
nothing
protects
centralization
more
than
a
good
moat
and
by
the
way
you
know,
a
a
state
is
basically
an
organization
with
the
moat
provided
by
an
army
right
at
far.
That's
what
it
boils
down.
F
E
B
B
E
So
in
our
world,
if
I
were
to
point
to
one
block
to
to
the
the
next
better
future,
is
that
assumption
of
scarcity
is
not
enough
for
everybody
and
yes,
that's
how
everybody
thinks,
but
we
really
should
be
considering
seriously
how
we
teach
people
an
abundance
mindset,
not
some
hippie
ideals.
I'm
talking
about
raw
productivity,
responsibility,
freedom.
E
E
E
Money
right,
so
that's
definitely
in
a
way
and
we're
saying
well.
We
still
believe
that
until
the
point
where
we
think
that
we
cannot
succeed
with
better
product
crowd
development
collaborative
design.
At
that
point,
we
might
turn
to
vc
capital,
which
I
doubt
because
I'm
pretty
convinced
that
this
is
going
to
work.
But
we
I'm.
E
D
E
E
H
Times,
vcs
yeah
yeah
look:
we've
raised
18.1
million
dollars.
We
just
raised
10
million
dollars
at
a
500
million
valuation.
It
is
laughably
easy.
The
trick
is
to
have
no
ideology
and
then
to
build
something,
that's
better
than
something
that
already
exists
right.
Investors
are
so
desperate
for
anything,
that's
even
slightly
better.
H
You
know
we
build
this
thing.
It's
open
source,
it's
legitimately
better
than
everything
else
on
the
market.
Consumer
reports
says
top
eight
ass
system
and
then,
like
you,
know
the
vcs
who
are
interested
in.
How
is
it
going
to
make
money?
You
don't
get
to
invest.
Sorry
bro,
you
miss
out.
Oh
I'm
missing
out
on
something.
Oh,
how
do
I
get
in?
How
do
I
get
in?
Oh
now
you
want
to
get
in
no,
no
you're,
not
the
right
type
of
person.
H
H
H
H
I
I
think
that
there's
so
much
capital
out
there
today
that,
if
you
can
tell
a
story
like
this,
it
just
you
know
shows
up.
If
you
tell
a
story
about
how,
like
you
know,
I
don't
know,
I
mean
like
it's
laughable
when
I
see
all
like
the
social
good
stuff
right,
the
investors
try
to
hide
behind
us,
like
oh
we're
doing
it
for
social
good,
I'm
like
no,
no,
no,
we
are
ruthlessly
capitalist
and
we're
going
to
give
our
stuff
away.
Why
are
you
going
to
give
it
away?
H
H
H
By
with
that
last
part
so
like,
if
you
can
give
something
away
right,
that
other
companies
were
planning
to
rent
seek
on
for
100
billion
dollars-
and
I
mean
actually
just
rent-
it
produce
no
more
value,
just
rent
seek
for
100
billion
dollars.
If
you
can
take
that
out
of
the
economy,
that
might
benefit
you
more
than
making
a
billion
dollars
personally
and
that's
if
you
can
get
people
to
start
believing
that
mentality.
H
H
B
H
H
So
we
have,
we
have
we
have
ada
software
right
now,
it's
on
github,
the
three
open
pilot,
it's
competitive
with
tesla
autopilot
and
better
than
everything
else,
and
there's
companies
that
are
trying
to
you
know
companies
that
have
raised
hundreds
of
millions
even
billions
in
some
cases
that
have
inferior
software
to
us
right
those
companies.
Well,
what
can
they
possibly
offer?
They
can
offer
nothing
right
and
they're
not
going
to
be
able
to
raise
rounds
in
the
future
if
investors
ever
start
actually
doing
any
sort
of
due
diligence
and
being
like
wait.
A.
D
H
H
Right
like
it's
free,
it's
going
to
be
impossible
to
compete,
and
I
think
this
is.
This
is
like
a
really
wonderful
thing
about
open
source.
It's
not
some
hippie
collaboration,
stuff,
it's
actually
a
competitive
bludgeon
that
you
can
smash
into
large
corporations
and
be
like
brah.
I
just
gave
your
away
for
free.
F
I
would,
I
would
push
back,
though,
against
saying
that
this
means
that
capitalism
is
a
zero-sum
game.
On
the
contrary,
it's
a
it's
a
positive
sun
game
and
that's
why
it's
powerful
right,
if
you
take
something
that
was
a
hundred
billion
dollar
business
and
completely
destroyed
by
making
something
better
free
right,
well,
yeah
the
owners
of
that
business
lost
their
100
billion
dollars,
but
everybody
on
balance
is
actually
better
off
right.
Yes
between
you
and
the
buyer,
you
have
gained
more
than
the
old
seller
has
lost,
so
it's
a
positive
sum
game.
F
H
H
H
Exactly
like,
I
do
not
think
that
I
think
that
in
theory,
if
you
truly
had
free
markets,
of
course,
capitalism
is
positive.
Some
game,
but
free
markets
don't
mean
unregulated
markets.
You
have
to
have
some
amount
of
preventing
some
certain
bad
effects
in
markets.
Now
I
think
the
regulations
are
actually
making
the
whole
thing
even
worse.
They're
actually
benefiting
the
bad
actors
more
than
good
actors,
but.
F
I
mean
in
fact
you
just
brought
up,
I
think,
a
really
a
crucial
point,
which
is
the
falling
right
and
it's
true
of
a
market.
It's
also
true
of
of
a
social
network
right,
which
is,
on
the
one
hand
you
want
as
much
freedom
as
possible,
like
say,
free
speech
on
social
media.
On
the
other
hand,
or-
and
you
know
you
want-
you
know
as
much
freedom
in
the
economy
and
for
people
to
do
whatever
they
want
on.
F
The
tension
between
decentralization
that
and
centralization
is
that
without
those
rules,
things
collapse,
so
you
know
they'd
be
centralized,
but
you're
back
in
the
stone
age,
which
is
not
good
right,
but
without
too
much.
You
know
with
rules
that
are
overly
strict,
then
or
overly
centralized
in
terms
of
deciding,
for
example,
what
is
misinformation
and
whatnot?
Then
you
actually
kill
the
golden
goose
right.
H
F
And
there
will
be
different,
you
know
you
know,
look
at
reddit
versus
twitter
right,
they
have
very
different
rules
and
they
and
therefore
they
have
very
different
roles
and
and
vice
versa.
So
the
answer
is
not
going
to
be
the
same
for
everybody,
but
I
think
in
any
platform
any
market
any
you
know
social
organization.
F
You
are
going
to
have
a
set
of
rules
that
have
to
be
somehow
enforced
by
someone
with
the
power
to
enforce
them,
like
the
benevolent
dictator
of
an
open
source
project
and
and
and
and
really
a
lot
of
the
problem.
The
research
problem,
if
you
will,
is
figuring
out
what
are
the
best
rules?
What
is
the
best
organization
they
have
for,
for
whatever
is
that
you
want
to
accomplish
sure.
That's.
H
You
know
we
have
a
nice
gps,
localizer
all
built
into
the
thing
that
I
think
can
definitely
be
reused
for
tons
of
other
applications.
We
think
of
openpilot
more
like
an
operating
system
for
robotics
than
just
a
self-driving
car.
You
know
the
product
we're
coming
out
with
next
month
is
the
common
body?
H
A
H
H
I
don't
have
slave
labor
right
like
who
will
dig
the
well
well,
I
get
a
well
digging
machine,
but
who
will
operate
the
well
digging
machine
right?
I
need
slaves
now.
You
know
I'm
not
a
believer
in
enslaving
humans,
as
most
people
today
aren't.
But
you
know
I'm
a
believer
in
enslaving
robots,
so
first
build
slaves
and
then
the
land
will
be
really
nice
and
I
was
gonna
make
my
bed
who's
to
wash
my
sheets.
Well,
the
robots.
Well.
E
D
D
H
E
B
E
B
H
B
H
Would
have
a
hard
time
getting
around
the
zero
marginal
cost
issue?
I'm
like
look
man
we're
going
to
make
open
file.
It's
going
to
be
10x,
better
everyone's
going
to
use
it
or
die
it's
going
to
be
like
android,
because
it
has
zero
marginal
cost.
You
can
make
a
10x
better
tractor,
but
that
doesn't
mean
everybody's
going
to
switch
because
there's
a
large
switching
cost.
That's
that's
not
just
that's
not
just
an
nre,
it's
recurring
expense
right.
They
have
to
redo
the
whole
supply
chain.
You
know,
just
just
like
the
machines
to
build.
H
Intelligence
to
sort
the
plastics
right,
yeah
yeah
and
that's
the
problem-
it's
intelligence,
recycling
in
general,
the
more
intelligence
you
can
apply
to
recycling
the
more
efficiently
you
can
do
it
right.
It's
very
easy
to
throw
an
iphone
in
a
bat
that
melts
out
all
the
precious
metals
it's
much
harder
to
reuse
the
screws,
because
you're
going
to
need
something
very
intelligent
to
remove
the
screws
without
damaging
them.
E
D
E
B
A
If
you
ever
want
to
go
sure
yeah
thanks,
I'm
curious.
If
you
were
so
so
the
listeners
who
who
are
are
listening
to
this,
they
might
be
thinking
all
this
sounds
cool.
But
what
can
I
as
a
listener?
Do
I'm
curious
if
you
had
any
advice
or
some
some
standard
person
listening
to
this?
What
type
of
things
would
you
tell
them?
You
might
want
to
think
about
learning,
x,
y
and
z,
or
go
into
x,
y
and
z?
What
would
your
recommendations
be.
E
E
D
D
E
D
E
F
Well,
I
the
first
thing
I
recommend
is-
and
you
know
not
necessarily
in
this
sword
but
like
figure
out
how
you
want
to
make
a
positive
difference
in
the
world
number
one
and
then
number
two
find
like-minded
people
right
because
doing
it
all
by
yourself.
You
know,
will
probably
with
the
impact
that
you
can
have
and
then
number
three
right
figure
out.
F
How
do
you
want
to
organize
yourselves
right
and
you
know
don't
be
constrained
by
the
you
know
by
the
existing
norms
right
there's
an
enormous
space
today
of
you
know
deciding
to
be
organized
in
different
ways,
for
the
thing
that
you
want
to
do
and-
and
you
know,
there's
no
recipe,
because
if
there
was
you
know,
people
will
be
doing
it
already,
but
there's
a
lot
of
you
know
other
examples
you
can
look
at.
Let
me
just
give
you
this.
F
F
But
I
think
this
is
the
way
this
is.
I
think
what
the
internet
makes
possible,
and
this
is,
I
think,
what
from
the
point
of
view
of
like
human
flourishing
is
most
desirable.
Is
that,
like
everybody,
finds
figures
out
how
they
want
to
impact
the
world,
they
find
the
like-minded
people,
and
then
they
organize
themselves
in
whatever
way
they
see
fit,
and
nobody
else
can
tell
them
not
to.
H
Go
for
it
go
on
github,
find
an
open
source
project.
You
like
submit
a
pull
request.
The
maintainer
will
tell
you
if
it's
bad,
if
it's
good,
they
will
be
excited
and
if
they
are
excited
and
get
it
merged,
go.
Do
this
again
do
this
for
four
years
and
you
will
know
more
about
how
to
program
than
any
graduate
of
any
cs
program
on
average.
H
Right,
like
you
know,
that's
what
I
love
about
the
internet.
It's
like
there's,
no
more
excuses,
go
out
there,
you
have
a
computer,
you
have
github
download,
vs
code,
it's
free
and
go
contribute
to
the
software,
and
then
you
know
do
this
for
four
years
and
you'll
find
that
you're
really
really
good.
A
H
F
F
F
You
know
europe
has
some
strengths,
but
I
don't
think
it
can
quite
compete
in
this
respect
and
I
think
if
you
look
at
china,
for
example,
in
many
ways
china
from
attack
point
of
view
is
a
wild
west,
but
in
many
ways
it
is
not
and
and
in
many
ways
is
actually
getting
worse
and
worse.
So
for
the
time
being,
I
think
it's
still
the
us.
In
fact,
what
I
would
say
is
that
we
are
still
too
dependent.
F
We
humanity
on
america.
We
would
be
much
safer
because
things
can
go
wrong
here
too,
and
they
are
going
wrong
with
things
like
walkism
and
and
whatnot,
and
so
we
are
too
dependent
on
america.
We'd
be
much
safer
if
we
had
half
a
dozen
americas
around
the
world
and
that's
one
with
hope.
You
know
china
would
be
an
europe
and
african
latin
america
and
whatnot
then,
and
maybe
it
will
happen,
although
assuredly
not
in
the
short
term.
E
And,
of
course,
open
source
can
help
spread.
The
word
people
ask
me
all
the
time.
Why
don't
I
build
tractors
in
africa
and
my
response
is
well
here.
We
have
the
stability
if,
if
we
can't
make
the
economic
model
work
here,
we're
not
going
to
make
it
work
in
africa
and
in
africa
it
would
be
much
harder
to
go
through
the
development
process
without
the
resources
available
like
here.
A
A
B
F
F
H
H
A
A
Excellent
well,
this
has
been
wonderful.
I
think
that
maybe
the
world
actually
will
have
changed
because
of
this
at
least
I'm
crossing
my
fingers,
you
can
just
just
tell
everyone
where
you
can
find
it
so
we'll
pedro
george
martin,
where
they
can
find
you
and
yeah.
Thank
you
so
much.
This
has
been
great
yeah.