►
From YouTube: Twitch Channel Introduction / 2019 Project review
Description
Watch this for a short introduction to this twitch channel, with details about what content I plan on streaming and why I'm doing it. Broadcasted live on Twitch -- Watch live at https://www.twitch.tv/rhyolight_
A
A
What's
gonna
be
on
this
channel,
what
I'm
gonna
use
it
for
and
the
content
that
you
can
expect
to
see
here
as
I
go
along,
but
first
I
need
to
put
some
music
on,
because
that's
what
people
tend
to
do
this
time,
that's
going
to
be
some
lot
to
a
stock
cake.
It
has
European
quintet
so
I'm
just
having
some
music
on
the
background
here.
A
A
Otherwise,
I
am
going
to
refer
to
my
white
board
here,
so
I'm
gonna.
Do
this
a
lot
I
think
this
is
my
camera,
so
I'm
gonna
do
this
a
lot
if
I
have
a
Content
that
I
want
to
talk
about
I'll
draw
a
lot
of
stuff
on
a
whiteboard.
This
is
just
sort
of
to
remind
me
what
to
talk
about.
So,
first
of
all,
my
name
is
Matt
Taylor
I
am
a
software
engineer
by
trade.
A
A
So
when
we
did
that
we
created
new
meta,
org
and
I
took
over
as
the
open
source,
community
manager
and
I
still
work
in
that
capacity
for
noventa,
so
I'm
told
you
anything
about
what
we
do
yet,
but
at
least
that's
my
role
at
noventa
I
actually
started
off
as
a
front-end
engineer,
but
now
I
manage
essentially
Numenta
org.
The
public
HTML
talk
about
HTM
in
a
bit.
A
So
if
you
want
more
info
about
Numenta,
noventa,
comm
or
the
public
facing
side
of
Numenta,
then
it's
a
dot
work.
So
if
I'll
be
talking
also
about
HTML
on
this
channel
is
not
going
to
be
entirely
about
HTM,
but
everything
I
do
essentially
is
related
to
HTM.
Htm
is
hierarchical
temporal
memory
and
it's
a
technology
that
Numenta
has
been
developing
since
it's
it
was
created
like
over
a
decade
ago.
Our
mission
is
to
study
how
intelligence
works
figure
out
how
intelligence
works
in
the
human
brain,
specifically
starting
with
the
neocortex.
A
That's
where
we've
always
looked
first
is
to
try
and
understand
the
core
algorithms
that
are
happening
in
the
new
cortex
and
then
apply
them
in
software,
essentially
try
and
create
them
and
non
replicate
them
a
non-biological
system
so
to
understand
intelligence
well
enough
in
the
brain
to
create
it
outside
the
brain.
That's
our
mission,
so
HTM
is
a
result
of
that.
It's
basically
sort
of
a
collective
term
for
all
the
things
we've
learned
up
until
this
point,
and
so
I'll
be
talking
about
different
aspects
of
HTM.
A
It's
a
it's
a
you
know
a
low-level
model
of
how
neurons
work
in
in
the
neocortex
talking
about
different
types
of
synapses
and
connections.
So
I'll
talk
about
that
stuff,
as
as
content
on
this
channel
as
it
go
along.
Let
me
show
you
my
schedule,
real
quick.
If
you,
if
you
go
to
my
twitch
page,
oh
man,
now
I
gotta
pause
this
boss,
okay,
so
this
is
basically
live
me
ten
seconds
ago.
So
this
schedule
the
way
I
I,
will
try
and
keep
updated
here.
A
I'm,
probably
not
gonna.
Do
the
work
session
9
a.m.
work
session
today,
because
I'm
sort
of
doing
this
right
now
and
I
and
I
might
who
knows
I
might
I
have
such
to
work
on,
but
I
definitely
will
will
keep
a
regular
1
p.m.
Monday
chat
open.
Where-
and
this
is
this
will
be
I'm
gonna
do
this
today
1.
So
this
will
be
the
first
episode
of
it
and
I'm
gonna
invite
you
know
anyone
to
come
and
talk
about
AI,
specifically
or
neuroscience.
A
You
can
talk
about
AG
I,
if
you
like,
although
that's
not
my
focus,
I,
really
like
to
focus
on
the
details,
the
implementations
of
things
and
we're
and
we're
far
enough
away
from
AGI
it's
hard
to
do
to
talk
about
the
details
or
implementations
of
things.
So
a
lot
of
what
I
talk
about
is
low
level
AI
stuff
and
the
neuroscience
that
sort
of
relates
to
the
low
level.
A
the
eye
eye
stuff
that
we
like
to
talk
about.
So
I
will
start
off
that
at
1
p.m.
A
today
specific
time
and
I'm
gonna
talk
about
AI
in
general
and
I'll
related
a
little
bit
to
game
development.
Since
twitch
is,
you
know
so
focused
on
gamers
and
most
of
the
other,
like
programmers
and
devs
that
I
found
out
there
are
game
developers.
So
I'll
talk
a
bit
about
how
AI
could
be
applied
in
game
development.
What
is
AI
week
versus
strong
AI
sort
of
different
classifications
of
AI?
What's
deep
learning?
What's
a
neural
network
or
it
is
a
biologically
realistic,
neural
network.
A
Spatial
versus
temporal
pattern,
recognition
that
type
of
stuff
so
I'll
talk
about
that
at
one
o'clock.
Tuesdays
and
Thursdays
are
sort
of
my
work
days,
so
I
will
have
a
big
work
session,
probably
Tuesday
9
a.m.
and
Thursday
at
9
a.m.
for.
However
long
I
want
to
go.
These
are
gonna,
be
I.
Don't
talk
about
the
projects
later,
I'll
tell
you
it's
like
whatever
project
I'm
working
on
Wednesday
I'm,
totally
offline,
because
I
Drive
to
the
Newman.
A
This
is
my
office
close
to
my
house
and
I
Drive
up
to
the
headquarters
up
in
Redwood
City
on
Wednesday,
so
I'll,
be
there
Wednesdays
and
I'm,
probably
gonna,
be
offline.
I
can't
promise
that
I'll
produce
any
content
on
Wednesdays
and
Fridays.
I've
got
a
question
mark
cuz,
I,
don't
know,
I,
don't
know
what
I'm
gonna
do
yet
I'll,
probably
maybe
do
something
we'll
see.
So
that
was
the
schedule.
Okay
projects.
A
A
First
off
the
HTM
videos,
I
make
educational
HTM
videos
and
you
can
call
them
HTM
school,
so
I
have.
For
example,
let's
see
oh
yeah,
just
click
on
in
the
about
me.
Just
click
here
and
you'll
see
the
YouTube
page,
it'll
start
saying
so
all
these
videos
here
these
have
tons
of
graphics
and
editing
and
stuff,
so
I
will
be
working
on
those
eyes
and
they're,
not
all
HTM
squat,
like
I'm
working
on
a
video
right.
A
Now,
that's
not
going
to
be
in
this
series,
but
it's
going
to
be
on
the
YouTube
page
and
it
I'm
putting
a
lot
of
time
and
effort
into
it.
So
so
that
involves
like
Adobe
Premiere,
After,
Effects,
Photoshop
and
and
for
graphics
and
then
I'll
do
and
then
editing
and
stuff
and
it'll
also
involve
potentially
recording
I.
A
If
you,
if
you
look
it's
particularly
like
topology
or
temporal
memory,
videos,
there's
always
this
3d,
that's
a
big
WebGL
animation
that
you
can
fly
through
and
see
you
like
the
HTM
system
running.
So
there's
a
lot
of
code
behind
that.
So
that
might
be
interesting
to
show
okay,
aside
from
no
wait
I'm
in
the
right
place.
Aside
from
that
aside
from
HTM
videos,
what's
the
next
thing,
I
could
be
working
on
our
community
projects
and
I
want
to
do
more
of
this.
A
A
So
here's
an
example
I
would
love
to
work
on
this.
This
project,
you
want
to
know
the
details,
there's
a
bunch
of
stuff
in
here
of
sort
of
how
I
see
it.
The
next
thing
I
want
to
do
is
maybe
write.
Some
tests,
try
and
figure
out
how
I
can
write
a
test
system,
that's
sort
of
agnostic
to
the
implementation
that
can
define,
object,
recognition,
tasks
as
if
this
first
is
a
very
low
level.
So
anyway,
so
here's
one
of
them.
This
would
be
an
HTM
project.
A
That's
involving
the
community,
so
I
wouldn't
be
working
on
this
alone.
I
would
be
trying
to
define
the
project
and
the
scope
of
it
and
work
on
it
together
with
other
people.
Another
one
that,
if
you,
if
you
look
at
the
last
video,
did
yesterday,
there's
this
yeah
HTM
path.
This
is
a
one
of
our
four
members.
Marty.
A
Did
this
cool,
C,
C++
I,
think
demo
of
he
created
an
HTM
infer,
STUV
all
and
then
then
he's
got
like
the
grid
cell
modules
working
properly
as
an
encoder,
so
he's
got
grid
cell
modules
in
coding
space
anyway.
Look
at
the
video!
If
you
want
to
see
it,
but
another
project
is
digging
into
this
because
I
don't
have
c++
person,
never
done
C++,
honestly,
never
written
all
C++,
but
it
would
be
an
interesting
community
thing
to
go
and
get
his
code
try
and
get
it
running
locally.
A
A
A
If
you're,
not
a
software
engineer,
you
might
not
care
about
this,
but
I.
We
new
mentis
platform
for
intelligent
computing
or
new
pick
has
been
around
for
a
long
time.
This
is
the
open
source
HTML
imitation
that
I'm
the
community
manager
for
and
so
I've
managed
this
project
for
years.
It
has
been
sort
of
in
maintenance
mode
for
a
couple
years
now
and
in
the
meantime,
and
it
runs
on
Python
2.7.
So
in
the
mean
time,
2.7
end-of-life
is
coming
soon,
so
it
2.7.
A
We
cannot
depend
on
anymore
after
January,
so
we
have
to
upgrade
new
pick
or
this
codebase
is
instantly
obsolete.
Essentially
so
one
of
the
things
I'll
be
working
on
is
I
would
call
this
like
a
legacy
software
engineering
problem.
Ok,
it's
taking
a
legacy
code
base
and
updating
it
to
a
new
environment,
Python
3.
So
this
is
probably
this
has
to
happen
this
year,
sometime
and
I'm
sure
I'm
going
to
be
in
the
in
the
throes
of
this.
A
A
What
we've
learned
so
far
from
the
brain
and
improve
the
performance
of
existing
systems,
so
I'll
have
I
will
be
working
on
this
quite
a
bit,
especially
with
by
trying
to
introduce
sparsity
it's
in
two
otherwise
dense
neural
networks,
because
we
know
sparsity
is
important
in
the
brain.
We
know
it
helps
improve
results
in
machine
learning,
but
the
way
HTM
introduces
spark.
So
here
the
concepts
of
spatial
pooling
when
those
are
applied
properly,
we
think
it
will
help
perform
machine
learning
systems
perform
better.
A
So
we've
already
get
we're
already.
You
have
a
paper,
we
have
almost
written
on
this
and
it
will
come
out
very
soon
I'm
going
show.
This
is
the
video
that
I'm
working
on
is
all
about.
This
so
you'll
see
me
working
on
this
quite
a
bit
this
year.
Hopefully
as
long
as
this
new
pictu
doesn't
take
too
long.
Okay.
A
You
know
so
that
is
essentially
the
the
channel
intro
and
the
project
review.
That's
what
I'm
be
working
on
for
this
whole
year,
I'm
gonna,
try
and
use
twitch
as
much
as
possible
and
just
stream
if
I'm
ever
doing,
engineering
work,
that's
relevant
to
any
one
of
these
things
that
I
think
someone
might
benefit
from
I'll,
just
stream
it
on
Twitch,
so
check
my
schedule.
A
If
you
want
to
know
the
schedule,
does
anyone
have
any
questions
and
see
if
there's
a
few
people
in
shadow,
even
where
to
help
you
more
than
happy
to
answer
any
questions
that
you
guys
have
this
music's
not
loud
enough
right,
I
think
it
just
went
off
okay.
This
is
my
opportunity
to
flip
sides
and,
if
you're
watching
this
and
you're
annoyed
by
the
turntable
of
this
I'm,
sorry
I
like
listening
to
records.
A
A
Okay,
so
it
doesn't
look
like
they're.
Any
questions
join
me
at
one
o'clock
today
and
I
want
to
talk
about
we're.
Gonna
talk
about
AI
I'm
gonna
talk
specifically
about
weak
versus
strong
AI.
What
is
today's
AI?
What
is
it
that's
doing
your
speech
recognition,
for
example,
when
you
talk
into
your
into
Siri
or
whatever,
what
type
of
AI
is
that,
what's
doing
your
photo
labeling
in
google
photos
or
facebook?
That's
telling
you
hey!
There's
your
friend
Pam.
A
What
type
of
AI
is
that
so
I'll
talk
about
that
and
we'll
talk
about
classic
AI
and
what
is
classic
AI
expert
systems,
for
example,
when
you
get
into
neural
networks
at
what
point
are
you
talking
about
deep
learning
versus
neural
network?
So
it
is
deep
learning,
a
neural
network
or
neural
network
at
deep
learning.
So
just
like
disambiguating
some
terms
and
then
I
would
also
like
to
talk
about
specifically
like
how
could
you
apply
AI
for
games
so
we'll
talk
about
that
at
1:00
p.m.