Latest / Omni Talk Retail / How Agentic AI Will Transform Retail Forever With David Dorf Of AWS | 5 Insightful Minutes

Transcript

How Agentic AI Will Transform Retail Forever With David Dorf Of AWS | 5 Insightful Minutes · Omni Talk Retail

• 0:00
  Foreigning
• 0:09
  us
• 0:10
  now
• 0:10
  for
• 0:10
  five
• 0:10
  insightful
• 0:11
  minutes
• 0:11
  is
• 0:11
  David
• 0:11
  Dorff.
• 0:12
  David
• 0:12
  is
• 0:13
  the
• 0:13
  head
• 0:13
  of
• 0:13
  Retail
• 0:14
  industry
• 0:14
  solutions
• 0:14
  at
• 0:15
  AWS.
• 0:16
  David,
• 0:16
  welcome
• 0:17
  back
• 0:17
  to
• 0:17
  Omnitalk.
• 0:18
  And
• 0:18
  let's
• 0:18
  get
• 0:18
  started
• 0:18
  with
• 0:19
  this.
• 0:20
  Gen
• 0:20
  AI
• 0:20
  was
• 0:21
  the
• 0:21
  big
• 0:21
  topic
• 0:21
  for
• 0:22
  the
• 0:22
  last
• 0:22
  two
• 0:22
  years,
• 0:22
  and
• 0:22
  now
• 0:23
  2025
• 0:24
  seems
• 0:24
  to
• 0:24
  be
• 0:24
  the
• 0:24
  year
• 0:24
  of
• 0:25
  the
• 0:25
  agent,
• 0:25
  or
• 0:26
  agentic
• 0:26
  AI
• 0:27
  as
• 0:27
  some
• 0:27
  might
• 0:27
  call
• 0:27
  it.
• 0:28
  How
• 0:28
  do
• 0:28
  you
• 0:29
  define,
• 0:29
  quote,
• 0:29
  unquote,
• 0:30
  agents?
• 0:31
  Ah,
• 0:32
  yes,
• 0:32
  good
• 0:33
  question.
• 0:34
  So
• 0:34
  genai
• 0:35
  is
• 0:35
  about
• 0:35
  creating
• 0:35
  content,
• 0:36
  while
• 0:36
  agents
• 0:37
  are
• 0:37
  really
• 0:37
  about
• 0:37
  taking
• 0:38
  action.
• 0:39
  They
• 0:39
  both
• 0:39
  use
• 0:39
  foundation
• 0:40
  models
• 0:40
  underneath,
• 0:41
  but
• 0:41
  there's
• 0:42
  really
• 0:42
  three
• 0:42
  key
• 0:42
  things
• 0:43
  for
• 0:43
  agents.
• 0:44
  Number
• 0:44
  one,
• 0:44
  they're
• 0:44
  autonomous,
• 0:45
  so
• 0:45
  they
• 0:45
  have
• 0:46
  a
• 0:46
  role
• 0:46
  and
• 0:46
  they
• 0:47
  require
• 0:47
  minimal
• 0:48
  human
• 0:48
  oversight.
• 0:49
  Number
• 0:49
  two,
• 0:50
  there's
• 0:50
  reasoning
• 0:50
  involved,
• 0:51
  so
• 0:51
  they
• 0:51
  actually
• 0:52
  do
• 0:52
  this
• 0:52
  thing
• 0:52
  called
• 0:52
  chain
• 0:53
  of
• 0:53
  thought
• 0:53
  where
• 0:53
  they
• 0:53
  break
• 0:54
  down
• 0:54
  problems
• 0:55
  into
• 0:55
  smaller
• 0:55
  steps.
• 0:56
  And
• 0:57
  number
• 0:57
  three
• 0:57
  is
• 0:57
  they
• 0:57
  typically
• 0:58
  have
• 0:58
  some
• 0:58
  sort
• 0:58
  of
• 0:59
  tools
• 0:59
  or
• 0:59
  access
• 1:00
  to
• 1:00
  data
• 1:00
  in
• 1:00
  tools
• 1:01
  to
• 1:01
  be
• 1:01
  able
• 1:01
  to
• 1:02
  take
• 1:02
  action
• 1:03
  to
• 1:03
  do
• 1:03
  something.
• 1:04
  So
• 1:04
  those
• 1:05
  three
• 1:05
  things
• 1:05
  are
• 1:06
  really
• 1:06
  what
• 1:06
  set
• 1:06
  agents
• 1:07
  apart.
• 1:07
  And
• 1:08
  for
• 1:08
  example,
• 1:08
  you
• 1:09
  can
• 1:09
  use
• 1:09
  generative
• 1:09
  AI
• 1:10
  to
• 1:10
  just
• 1:10
  create
• 1:11
  an
• 1:11
  image,
• 1:11
  but
• 1:12
  an
• 1:12
  agent,
• 1:12
  you
• 1:13
  could
• 1:13
  say,
• 1:13
  I
• 1:13
  need
• 1:13
  a
• 1:14
  specific
• 1:14
  image
• 1:15
  that
• 1:15
  is
• 1:16
  relaxing
• 1:16
  and
• 1:17
  I
• 1:17
  want
• 1:17
  it
• 1:17
  posted
• 1:17
  to
• 1:18
  a
• 1:18
  website.
• 1:18
  And
• 1:18
  it
• 1:19
  can
• 1:19
  do
• 1:19
  all
• 1:19
  of
• 1:19
  that.
• 1:19
  Right.
• 1:20
  It's
• 1:20
  taking
• 1:20
  multiple
• 1:21
  steps
• 1:21
  and
• 1:21
  it's
• 1:21
  doing
• 1:22
  action.
• 1:23
  So
• 1:23
  there
• 1:23
  are
• 1:23
  lots
• 1:23
  of
• 1:24
  different
• 1:24
  genai
• 1:24
  use
• 1:25
  cases
• 1:25
  out
• 1:25
  there,
• 1:25
  but
• 1:25
  agentic
• 1:26
  genai
• 1:27
  is
• 1:27
  really
• 1:27
  where
• 1:28
  things
• 1:28
  are
• 1:28
  starting
• 1:28
  to
• 1:28
  explode.
• 1:30
  So,
• 1:30
  David,
• 1:30
  are
• 1:31
  retailers
• 1:31
  actually
• 1:32
  using
• 1:32
  these
• 1:33
  agents
• 1:33
  already
• 1:34
  today?
• 1:34
  And
• 1:35
  if
• 1:35
  so,
• 1:36
  how
• 1:36
  are
• 1:36
  they
• 1:36
  using
• 1:36
  them?
• 1:37
  Yeah,
• 1:38
  it's
• 1:38
  not
• 1:38
  quite
• 1:39
  as
• 1:39
  prevalent
• 1:39
  right
• 1:39
  now
• 1:39
  because
• 1:40
  it
• 1:40
  is
• 1:40
  pretty
• 1:40
  new,
• 1:41
  but
• 1:42
  for
• 1:42
  sure.
• 1:42
  So
• 1:43
  a
• 1:43
  great
• 1:44
  example
• 1:44
  is
• 1:44
  Amazon
• 1:45
  uses
• 1:45
  agents
• 1:46
  to
• 1:46
  do
• 1:46
  a
• 1:46
  lot
• 1:46
  of
• 1:47
  Java
• 1:47
  upgrades.
• 1:48
  Sounds
• 1:48
  like
• 1:48
  a
• 1:48
  boring
• 1:49
  task,
• 1:49
  but
• 1:49
  we
• 1:49
  have
• 1:50
  like
• 1:50
  30,000
• 1:51
  Java
• 1:51
  applications
• 1:52
  that
• 1:52
  need
• 1:52
  to
• 1:53
  be
• 1:53
  upgraded
• 1:53
  from
• 1:54
  an
• 1:54
  older
• 1:54
  version
• 1:54
  to
• 1:55
  a
• 1:55
  newer
• 1:55
  version.
• 1:56
  And
• 1:56
  we
• 1:56
  have
• 1:56
  agents
• 1:57
  that
• 1:57
  go
• 1:57
  through
• 1:57
  and
• 1:57
  do
• 1:57
  that
• 1:57
  and
• 1:58
  saved
• 1:58
  us
• 1:58
  about
• 1:58
  $260
• 1:59
  million
• 2:00
  last
• 2:00
  year.
• 2:01
  Wow.
• 2:01
  And
• 2:02
  in
• 2:02
  addition
• 2:02
  to
• 2:02
  just
• 2:03
  doing
• 2:03
  those
• 2:03
  upgrades,
• 2:03
  they'll
• 2:04
  also
• 2:04
  build
• 2:05
  unit
• 2:05
  tests
• 2:06
  and
• 2:06
  they'll
• 2:06
  do
• 2:06
  documentation
• 2:07
  and
• 2:07
  things
• 2:08
  like
• 2:08
  that.
• 2:08
  And
• 2:09
  we've
• 2:09
  got
• 2:09
  some
• 2:09
  other
• 2:09
  retailers
• 2:10
  doing
• 2:10
  some
• 2:10
  cool
• 2:10
  stuff
• 2:11
  like
• 2:11
  Tapestry,
• 2:11
  the
• 2:12
  luxury
• 2:12
  retailer.
• 2:13
  They
• 2:14
  actually
• 2:14
  have
• 2:14
  an
• 2:15
  agent
• 2:15
  that
• 2:16
  helps
• 2:16
  find
• 2:17
  data
• 2:17
  and
• 2:17
  answer
• 2:18
  questions.
• 2:18
  It
• 2:19
  will
• 2:19
  take
• 2:19
  your
• 2:20
  textual
• 2:21
  question,
• 2:21
  figure
• 2:22
  out
• 2:22
  what
• 2:22
  database
• 2:22
  it
• 2:23
  needs
• 2:23
  to
• 2:23
  answer
• 2:23
  that,
• 2:24
  convert
• 2:24
  it
• 2:24
  to
• 2:24
  SQL,
• 2:25
  do
• 2:25
  the
• 2:25
  SQL
• 2:25
  query,
• 2:26
  and
• 2:26
  come
• 2:26
  back
• 2:26
  with
• 2:27
  it
• 2:27
  with
• 2:27
  an
• 2:27
  English
• 2:28
  answer.
• 2:29
  So
• 2:29
  it's
• 2:29
  kind
• 2:29
  of
• 2:29
  like
• 2:30
  an
• 2:30
  analyst,
• 2:30
  but
• 2:31
  it
• 2:31
  gives
• 2:31
  the
• 2:31
  business
• 2:32
  a
• 2:32
  chance
• 2:32
  to
• 2:32
  be
• 2:33
  able
• 2:33
  to
• 2:33
  get
• 2:33
  this
• 2:33
  information
• 2:34
  without
• 2:34
  necessarily
• 2:35
  having
• 2:35
  to
• 2:36
  create
• 2:36
  all
• 2:36
  sorts
• 2:36
  of
• 2:37
  weird
• 2:37
  SQL
• 2:37
  queries.
• 2:39
  And
• 2:39
  then
• 2:39
  we
• 2:39
  have
• 2:40
  a
• 2:40
  tire
• 2:40
  retailer
• 2:42
  who's
• 2:42
  giving
• 2:42
  recommendations
• 2:44
  for
• 2:44
  tires.
• 2:45
  So
• 2:45
  You.
• 2:45
  You,
• 2:46
  for
• 2:46
  example,
• 2:46
  can
• 2:47
  say
• 2:47
  what
• 2:47
  make
• 2:48
  and
• 2:48
  model
• 2:48
  you
• 2:48
  have
• 2:48
  and
• 2:49
  what
• 2:49
  you
• 2:49
  use
• 2:49
  your
• 2:49
  car
• 2:50
  for.
• 2:50
  And
• 2:50
  it
• 2:50
  goes
• 2:50
  and
• 2:50
  hits
• 2:51
  several
• 2:51
  different
• 2:51
  resources
• 2:52
  to
• 2:52
  figure
• 2:52
  out
• 2:52
  what
• 2:53
  the
• 2:53
  best
• 2:53
  tire
• 2:54
  recommendation
• 2:54
  might
• 2:55
  be.
• 2:55
  And
• 2:55
  it
• 2:55
  kind
• 2:55
  of
• 2:56
  gives
• 2:56
  you
• 2:56
  the
• 2:56
  reasoning
• 2:57
  behind
• 2:57
  why
• 2:57
  it
• 2:57
  recommends
• 2:58
  one
• 2:58
  tire
• 2:59
  over
• 2:59
  another,
• 2:59
  which
• 3:00
  I
• 3:00
  think
• 3:00
  is
• 3:00
  pretty
• 3:00
  unique.
• 3:02
  Wow.
• 3:03
  And
• 3:03
  I
• 3:03
  don't
• 3:03
  know
• 3:03
  about
• 3:03
  you,
• 3:03
  but
• 3:03
  I
• 3:04
  had
• 3:04
  no
• 3:04
  idea
• 3:04
  the
• 3:05
  agents
• 3:05
  were
• 3:05
  already
• 3:05
  being
• 3:05
  deployed
• 3:06
  to
• 3:06
  the
• 3:07
  degree
• 3:07
  which
• 3:07
  David
• 3:07
  was
• 3:07
  saying.
• 3:08
  I
• 3:08
  thought
• 3:08
  this
• 3:08
  was
• 3:08
  just
• 3:08
  kind
• 3:08
  of
• 3:09
  the
• 3:09
  new
• 3:09
  buzzword
• 3:10
  idea
• 3:11
  for
• 3:11
  2025,
• 3:12
  but.
• 3:12
  So
• 3:12
  with
• 3:12
  that
• 3:12
  said,
• 3:12
  David,
• 3:14
  I'm
• 3:14
  curious.
• 3:14
  How
• 3:14
  do
• 3:15
  you
• 3:15
  see
• 3:15
  agents
• 3:16
  changing
• 3:17
  retail
• 3:17
  even
• 3:18
  more
• 3:18
  over
• 3:18
  the
• 3:19
  next,
• 3:19
  like,
• 3:19
  five
• 3:20
  years?
• 3:20
  Let's
• 3:20
  say,
• 3:20
  what
• 3:21
  are
• 3:21
  the
• 3:21
  big
• 3:21
  things
• 3:22
  that
• 3:22
  are
• 3:22
  still
• 3:22
  going
• 3:22
  to
• 3:23
  change
• 3:23
  on
• 3:23
  this
• 3:23
  front?
• 3:24
  This
• 3:25
  is
• 3:25
  where
• 3:25
  things
• 3:25
  can
• 3:25
  get
• 3:25
  a
• 3:26
  little
• 3:26
  bit
• 3:26
  crazy.
• 3:27
  One
• 3:27
  type
• 3:28
  of
• 3:28
  agent
• 3:28
  is
• 3:28
  a
• 3:28
  computer
• 3:29
  use
• 3:29
  agent.
• 3:30
  So
• 3:30
  those
• 3:31
  were
• 3:31
  announced
• 3:31
  by
• 3:31
  Anthropic
• 3:32
  and
• 3:32
  OpenAI.
• 3:33
  And
• 3:33
  basically
• 3:34
  it
• 3:34
  has
• 3:34
  control
• 3:35
  of
• 3:35
  your
• 3:35
  browser
• 3:36
  and
• 3:36
  so
• 3:36
  it
• 3:36
  takes
• 3:36
  a
• 3:37
  picture
• 3:37
  of
• 3:37
  the
• 3:37
  browser
• 3:37
  screen
• 3:38
  so
• 3:38
  it
• 3:38
  knows
• 3:38
  where
• 3:38
  things
• 3:39
  are
• 3:39
  and
• 3:39
  it
• 3:39
  can
• 3:39
  navigate
• 3:40
  with
• 3:40
  the
• 3:40
  mouse
• 3:40
  and
• 3:40
  keyboard.
• 3:41
  And
• 3:41
  the.
• 3:42
  The
• 3:42
  demonstration
• 3:43
  that
• 3:43
  OpenAI
• 3:44
  gave
• 3:44
  was
• 3:44
  using
• 3:45
  Instacart
• 3:45
  to
• 3:46
  do
• 3:46
  your
• 3:46
  grocery
• 3:46
  shopping.
• 3:47
  There
• 3:47
  was
• 3:47
  even
• 3:47
  a
• 3:47
  New
• 3:48
  York
• 3:48
  Times
• 3:48
  article
• 3:48
  where
• 3:49
  the
• 3:49
  author
• 3:49
  bought
• 3:50
  eggs
• 3:50
  using
• 3:51
  a
• 3:51
  computer
• 3:51
  agent,
• 3:51
  just
• 3:52
  saying,
• 3:52
  go
• 3:52
  find
• 3:52
  me
• 3:52
  eggs.
• 3:53
  So
• 3:54
  if
• 3:54
  you
• 3:54
  think
• 3:54
  about
• 3:54
  that
• 3:55
  for
• 3:55
  a
• 3:55
  minute,
• 3:55
  how
• 3:56
  does
• 3:56
  that
• 3:56
  really
• 3:56
  upend
• 3:57
  the
• 3:57
  retail
• 3:57
  industry
• 3:58
  with
• 3:58
  shopping?
• 3:58
  If
• 3:59
  people
• 3:59
  are
• 3:59
  using
• 4:00
  agents
• 4:00
  to
• 4:01
  do
• 4:01
  the
• 4:01
  shopping
• 4:01
  for
• 4:01
  them
• 4:02
  now,
• 4:02
  they're
• 4:02
  not
• 4:03
  going
• 4:03
  to
• 4:03
  do
• 4:03
  it
• 4:03
  for,
• 4:03
  like,
• 4:03
  fashion,
• 4:04
  but
• 4:04
  for
• 4:04
  replenishable
• 4:05
  items,
• 4:05
  grocery,
• 4:06
  that
• 4:06
  sort
• 4:06
  of
• 4:06
  thing.
• 4:06
  I
• 4:07
  can
• 4:07
  see
• 4:07
  that
• 4:07
  happening
• 4:08
  today.
• 4:09
  The
• 4:09
  user
• 4:10
  interface
• 4:10
  is
• 4:10
  made
• 4:10
  for
• 4:10
  humans
• 4:11
  and
• 4:11
  it's
• 4:11
  cluttered.
• 4:12
  And
• 4:12
  maybe
• 4:13
  we
• 4:13
  need
• 4:13
  to
• 4:13
  optimize
• 4:14
  that
• 4:14
  for
• 4:14
  agents
• 4:15
  just
• 4:15
  like
• 4:16
  we
• 4:16
  did
• 4:16
  for
• 4:16
  mobile.
• 4:17
  We
• 4:17
  have
• 4:17
  mobile
• 4:17
  optimized
• 4:18
  sites
• 4:18
  now.
• 4:18
  Maybe
• 4:19
  we'll
• 4:19
  have
• 4:19
  agent
• 4:19
  optimized
• 4:20
  sites
• 4:20
  that
• 4:20
  make
• 4:20
  it
• 4:20
  easier
• 4:21
  for
• 4:21
  agents
• 4:21
  to
• 4:22
  buy
• 4:22
  things
• 4:22
  on
• 4:22
  people's
• 4:23
  behalf.
• 4:23
  And
• 4:24
  then
• 4:24
  you
• 4:24
  have
• 4:24
  to
• 4:24
  start
• 4:24
  to
• 4:25
  think
• 4:25
  about
• 4:25
  what
• 4:25
  are
• 4:25
  some
• 4:26
  of
• 4:26
  the
• 4:26
  other
• 4:26
  things
• 4:26
  that
• 4:26
  could
• 4:27
  happen?
• 4:27
  What
• 4:27
  happens
• 4:28
  with
• 4:28
  advertising?
• 4:29
  There's
• 4:30
  no
• 4:30
  use
• 4:30
  in
• 4:30
  advertising
• 4:31
  to
• 4:31
  an
• 4:31
  agent.
• 4:32
  Should
• 4:32
  I
• 4:33
  be
• 4:33
  personalizing?
• 4:34
  Maybe
• 4:35
  that
• 4:35
  doesn't
• 4:35
  make
• 4:35
  sense.
• 4:36
  How
• 4:36
  do
• 4:36
  I
• 4:36
  affect
• 4:37
  a
• 4:37
  person
• 4:38
  when
• 4:38
  they're
• 4:38
  using
• 4:38
  an
• 4:38
  agent?
• 4:39
  Through
• 4:39
  things
• 4:39
  like
• 4:39
  promotions.
• 4:42
  An
• 4:42
  agent
• 4:42
  doesn't
• 4:42
  care
• 4:43
  about
• 4:43
  a
• 4:43
  promotion.
• 4:43
  So
• 4:44
  how
• 4:44
  do
• 4:44
  I
• 4:44
  get
• 4:44
  that
• 4:44
  information
• 4:45
  to
• 4:45
  a
• 4:45
  person?
• 4:46
  What
• 4:46
  about
• 4:46
  loyalty?
• 4:47
  That
• 4:48
  kind
• 4:48
  of
• 4:48
  gets
• 4:48
  thrown
• 4:49
  up
• 4:49
  if
• 4:49
  I
• 4:49
  can't
• 4:49
  really
• 4:50
  affect
• 4:50
  behavior.
• 4:51
  And
• 4:52
  maybe
• 4:52
  something
• 4:52
  like
• 4:53
  even
• 4:53
  search.
• 4:54
  We
• 4:54
  have
• 4:55
  a
• 4:55
  whole
• 4:55
  industry
• 4:55
  around
• 4:55
  search
• 4:56
  engine
• 4:56
  optimization.
• 4:57
  What
• 4:57
  if
• 4:57
  it's
• 4:57
  agents
• 4:57
  that
• 4:58
  are
• 4:58
  out
• 4:58
  there
• 4:58
  doing
• 4:58
  the
• 4:59
  shopping
• 4:59
  for
• 4:59
  us?
• 5:00
  How
• 5:00
  is
• 5:01
  search
• 5:01
  engine
• 5:01
  optimization
• 5:02
  going
• 5:02
  to
• 5:02
  change
• 5:03
  and
• 5:03
  how
• 5:03
  do
• 5:03
  we
• 5:03
  influence
• 5:04
  things
• 5:04
  differently?
• 5:05
  I
• 5:05
  think
• 5:05
  we're
• 5:05
  on
• 5:06
  the
• 5:06
  cusp
• 5:06
  of
• 5:06
  some
• 5:07
  really
• 5:07
  interesting
• 5:07
  changes
• 5:08
  to
• 5:08
  the
• 5:08
  retail
• 5:09
  industry
• 5:09
  that
• 5:10
  agents
• 5:10
  could
• 5:10
  bring
• 5:10
  us
• 5:11
  and
• 5:11
  we
• 5:11
  need
• 5:11
  to
• 5:11
  start
• 5:11
  thinking
• 5:12
  strategically
• 5:13
  about
• 5:13
  what
• 5:13
  are
• 5:13
  the
• 5:13
  ramifications
• 5:14
  here
• 5:14
  and
• 5:14
  how
• 5:15
  can
• 5:15
  we
• 5:15
  prepare.
• 5:16
  Well,
• 5:17
  David,
• 5:17
  you
• 5:17
  just
• 5:17
  blown
• 5:18
  both
• 5:18
  of
• 5:18
  our
• 5:18
  minds
• 5:19
  here.
• 5:22
  Unfortunately
• 5:23
  we
• 5:23
  have
• 5:23
  to
• 5:23
  get
• 5:24
  you
• 5:24
  out
• 5:24
  of
• 5:24
  here
• 5:24
  on
• 5:24
  this.
• 5:25
  But
• 5:25
  we
• 5:25
  will
• 5:26
  be
• 5:26
  thinking
• 5:26
  of
• 5:26
  many
• 5:26
  more
• 5:27
  questions.
• 5:28
  Where
• 5:29
  do
• 5:29
  people
• 5:29
  start?
• 5:29
  There's
• 5:30
  a
• 5:30
  lot
• 5:30
  that's
• 5:30
  changing.
• 5:31
  What
• 5:31
  do
• 5:31
  you
• 5:31
  recommend
• 5:32
  that
• 5:32
  retailers
• 5:33
  do
• 5:33
  now
• 5:33
  if
• 5:34
  they're
• 5:34
  wanting
• 5:35
  to
• 5:35
  explore
• 5:35
  taking
• 5:36
  advantage
• 5:36
  of
• 5:36
  gentic
• 5:37
  AI
• 5:37
  or
• 5:38
  just
• 5:38
  to
• 5:38
  kind
• 5:38
  of
• 5:38
  tackle
• 5:39
  some
• 5:39
  of
• 5:39
  the
• 5:39
  things
• 5:39
  that
• 5:39
  are
• 5:40
  going
• 5:40
  to
• 5:40
  be
• 5:40
  happening
• 5:40
  in
• 5:41
  the
• 5:41
  future?
• 5:42
  Yeah,
• 5:42
  yeah.
• 5:42
  So
• 5:43
  number
• 5:43
  one
• 5:43
  is
• 5:44
  basic
• 5:44
  blocking
• 5:44
  and
• 5:44
  tackling.
• 5:45
  Right.
• 5:45
  So
• 5:46
  don't
• 5:46
  get
• 5:46
  too
• 5:46
  caught
• 5:46
  up
• 5:47
  in
• 5:47
  gen
• 5:47
  hype.
• 5:48
  There's
• 5:48
  a
• 5:48
  lot
• 5:49
  of
• 5:49
  it
• 5:49
  out
• 5:49
  there.
• 5:49
  Number
• 5:50
  two,
• 5:50
  there's
• 5:50
  a
• 5:50
  lot
• 5:51
  of
• 5:51
  basic
• 5:51
  gen
• 5:51
  AI
• 5:52
  use
• 5:52
  cases
• 5:52
  that
• 5:52
  you
• 5:52
  can
• 5:53
  get
• 5:53
  a
• 5:53
  real
• 5:53
  good
• 5:54
  return
• 5:54
  on
• 5:54
  investment
• 5:55
  in
• 5:55
  things
• 5:56
  like
• 5:56
  creating
• 5:56
  product
• 5:56
  descriptions,
• 5:58
  definitely
• 5:59
  being
• 5:59
  able
• 5:59
  to
• 5:59
  do
• 6:00
  things
• 6:00
  like
• 6:00
  shopping
• 6:00
  assistants
• 6:01
  like
• 6:01
  Rufus.
• 6:02
  Those
• 6:02
  are
• 6:02
  some
• 6:02
  low
• 6:03
  hanging
• 6:03
  fruit
• 6:03
  that
• 6:03
  retailers
• 6:04
  should
• 6:04
  look
• 6:04
  at.
• 6:04
  And
• 6:05
  then
• 6:05
  we
• 6:05
  can
• 6:05
  start
• 6:05
  to
• 6:05
  get
• 6:05
  into
• 6:06
  the
• 6:06
  agentic
• 6:07
  areas
• 6:07
  where
• 6:07
  you've
• 6:08
  got
• 6:08
  processes
• 6:09
  that
• 6:09
  can
• 6:09
  be
• 6:09
  automated
• 6:10
  and
• 6:10
  you
• 6:10
  can
• 6:11
  really
• 6:11
  start
• 6:11
  to,
• 6:12
  I
• 6:12
  don't,
• 6:12
  I
• 6:13
  don't
• 6:13
  want
• 6:13
  to
• 6:13
  say
• 6:13
  replace
• 6:13
  people
• 6:14
  because
• 6:14
  we're
• 6:14
  not
• 6:14
  ready
• 6:14
  to
• 6:15
  do
• 6:15
  that
• 6:15
  yet,
• 6:15
  but
• 6:15
  really
• 6:16
  enhance
• 6:16
  the
• 6:16
  things
• 6:17
  that
• 6:17
  people
• 6:17
  are
• 6:17
  doing
• 6:18
  to
• 6:18
  be
• 6:18
  more
• 6:18
  efficient.
• 6:19
  Great
• 6:19
  stuff,
• 6:19
  David.
• 6:20
  Thank
• 6:20
  you.