►
From YouTube: Observing close and far objects across Visual Cortex
Description
How all level of visual hierarchy are doing object recognition on sensory input and informing each other via lateral connections, even without the hierarchical connections. An explanation of the Thousand Brains Theory of Intelligence.
A
A
A
I
I
tried
to
avoid
programming
at
this
point
in
my
career,
but
it
keeps
coming
back.
I
keep
having
to
do
it,
which
makes
it
frustrating
because
it
when
I
do
when
it
comes
back,
I
have
to
do
it.
I
have
to
like
get
up
to
speed,
sharpen
all
the
tools.
What's
the
new
build
platform,
what's
the
new,
it's
like
it's
so
much
harder
than
it
used
to
be
honestly.
A
A
Most
hierarchies
aren't
evident
like
most
hierarchies.
You
have
to
understand
the
object
and
how
it
relates
to
everything
for
it
to
be
evident,
but
then
there
are,
but
that
but
said,
I
don't
know
it.
This
is
it's
a
hard
topic,
but,
like
I
said,
you
can
pick
anything
in
the
world
whether
it's
a
neuron
or
an
object
in
the
room
and
find
hierarchies
associated
with
it
in
the
neuron.
If
you
look
at
all
of
its
connections,
you
can
follow
them
all
physically
right
and
discover
that
it's
a
part
of
some
hierarchy
in
a
network.
A
You
can
see
its
position
in
the
hierarchy,
but
nothing
really
else
is
right.
The
same
thing
with
an
object
in
in
a
room.
If
you
just
think
about
anything
that
this,
this
object
is
composed
of
other
objects
which
are
composed
of
other
objects,
and
it
is
a
is
a
part
of
bigger
composers.
You
can.
We
can
look
out
and
this
this
cup
is
a
part
of
the
room
that
I'm
in
the
room
is
a
part
of
the
building.
A
There's
like
us,
there's
a
everything
is
a
part
of
some
hierarchy
and
I'm,
not
sure
that
that
hierarchy
and
this
the
cell
hierarchy,
the
physical
hierarchy
that
you
can
look
at
it.
The
cell
are
much
different.
If
you
know
anything,
it's
kind
of
weird
take
on
it.
I
think,
but
I
haven't
figured
out
how
I
don't
know
that
there's
two
things
to
discover
here,
I
do
think
hierarchy
and
scale
are
super
related
because
scale.
A
That
Jeff
told
me
that
really
got
me.
So
if
you
are
I,
don't
know,
I
put
this
on
a
loop,
but
I,
don't
know
what
it's
gonna
get
where
it's
gonna
go.
Guess
it
does
over
here
so
say,
you're
at
on
Safari,
and
you
see
an
elephant
on
the
horizon
like
way
out
there.
You
see
an
elephant
and
it's
like
walking
across
the
plains,
and
you
see
its
silhouette
or
something.
A
If
you
took
us
straw
and
you
looked
through
the
straw
at
that
elephant
on
the
horizon,
you
could
still
tell
that
it's
an
elephant
right
because
you
can
see
in
its
profile.
It's
got
tusks,
it's
got
a
trunk,
it's
got
four
legs,
some
floppy
ears
or
something
you
can
tell
that
it's
an
elephant
just
with
that
little
straw
that
you
get
to
v-12
a
cortical
column
in
v1.
A
High
detail,
if
you
get
the
most
detail
in
that
area
of
your
field
of
view
in
visual
cortex
and
but
now
imagine
that
you've
walked
all
the
way
up
to
the
elephant,
and
it's
right
in
front
of
you
and,
and
you
don't
know,
let's
say
you're
in
you're
in
Safari,
but
you've.
Just
someone
spun
you
around
and
put
a
blindfold
on.
You
put
you
right
in
front
of
the
elephant
and
then
they
take
the
blindfold
off.
But
all
you
see
is
this
flat
elephant
and
you
don't
know
it's
an
elephant.
A
You
just
see
some
gray
wall,
wrinkly
wall
right
movie,
one,
let's
talk
about
just
a
v1,
just
if
you
want
just
v1,
is
gonna,
get
an
extremely
detailed
picture
of
the
super
close-up
view
of
elephant
skin
and
it's
gonna
be
confusing,
initially
I'm,
not
sure
that
you
right
away
know
from
just
v1
that
this
was
an
elephant
just
from
that
it
would
be
confusing.
It's
basically
see
a
texture,
some
type
of
weird
greyish,
colored
texture,
so
v1
would
not
be
be
able
to
identify
it.
A
You
can
imagine
it
not
being
able
to
identify
the
elephant
at
that
distance,
but
v3.
Let's
go
up
the
hierarchy
right
as
you
go
up
the
hierarchy,
these
the
higher
layers
levels.
Excuse
me
the
higher
levels
of
the
hierarchy.
You
still
receive
direct
sensory
input
up
through
the
thalamus
and
they
receive
it
in
a
larger
field
of
view,
so
they
get
less
detail
over
a
large
field
of
view.
This
is
where
scale
comes
in
the
small.
A
You
know
with
legs
on
a
safari:
they're
gonna
be
a
few
things,
so
you're
at
your
higher
levels
of
the
hierarchy
are
doing
object,
recognition
with
a
bigger
field
of
view
of
the
input,
but
let's
detail
it
doesn't
have
to
have
all
the
detail.
It's
putting.
It
has
a
different,
it's
doing
the
same
job
with
a
different
set
of
inputs,
and
so
it
knowing
that
it's
an
elephant,
can
then
through
lateral
connections
throughout
the
cortex,
inform
other
cortical
columns
that
eventually
get
to
I.
A
Don't
know
exactly
how
this
works
across
the
borders
or
whatever,
but
v1
gets
informed
of
what
v1
thinks
objects
are,
and
so
it
can
immediately
know
the
context
or
the
object.
That
is
looking
for
the
reference
frame,
the
object
that
it's
inspecting
so
that
it
can
add
detail
it
can
know
this
is
an
elephant,
and
that
makes
sense
it
can
confirm
it.
I'm
gonna
go
yeah
gray,
a
little
bit
of
hair,
it
can
confirm
it,
it
doesn't
and-
and
then
that's
not
even
talking
about
Sakata
around
you
know,
that's
that's
the
same
thing.
A
Your
v1
has
a
model
of
an
elephant
and
it
can
discuss
ikot
around
the
elephant.
It
can
get
lots
of
detail
about
its
corners.
It
can
get
candy
Dickon.
It
has
a
complete
model
of
an
elephant,
whether
you
see
it
at
a
distance
through
a
straw
or
it's
looking
right
up
front
to
Pat
it
right.
That's
the
ones.
That's
the
wavy.
One
sees
the
world
the
way
v3
sees
the
world
is
much
bigger
field
of
view,
much
less
detail,
but
it's
doing
the
same
thing.
It's
like
what
it.
What
is
this
thing?