00:00:05
Hey, everyone, welcome to The Huddle.
00:00:06
It's really great to be with you again.
00:00:09
It's been a little while,
00:00:10
but, we have a really cool guest today.
00:00:12
This is Azita Martin,
00:00:14
and I wanna talk about her in just a second.
00:00:16
But, the subject today is artificial intelligence -
00:00:21
all of the new capabilities, products, the things
00:00:24
that we're seeing around the world that are changing AI,
00:00:28
and AI is changing so much of what all of us do
00:00:31
and what we will do in the future.
00:00:33
I thought that it would be a great time to have Azita here.
00:00:35
Azita's an executive at Nvidia, who is,
00:00:38
all over the space and AI and chips
00:00:41
and started in gaming
00:00:42
and has worked their way all the way to what we do today,
00:00:45
and we'll, we'll get to that as well.
00:00:46
But Azita first, thank you for doing this.
00:00:49
- Of course.
- We got to speak
00:00:50
to an audience together in January
00:00:52
in New York at the National Retail Federation big show.
00:00:55
And after doing that, it went so well,
00:00:57
I thought the the best next step would be for you
00:00:59
to come here and talk to all
00:01:01
of us about the things you talked about while
00:01:03
we were in New York together.
00:01:04
- Absolutely. Happy to be here.
00:01:05
- Welcome to Bentonville.
- Thank you.
- Good to have you here.
00:01:08
First, let's start with you. Talk about who you are,
00:01:11
your career, and how you got from growing up,
00:01:14
and I know you'll say where, where you started
00:01:16
and how you got into artificial intelligence.
00:01:18
Yeah, absolutely.
00:01:20
I actually lived in a lot
00:01:22
of different places in the world,
00:01:23
but you know, went to high school in
00:01:26
a beautiful island in Spain, in Myorca.
00:01:29
And, you know, as a kid,
00:01:31
I was really good at math and physics,
00:01:34
and my father was an engineer, so the natural thing for me
00:01:37
to do was to become an engineer.
00:01:40
So, I came to the US at the age of 17 to go
00:01:43
to college and became an aerospace engineer
00:01:46
and actually worked at McDonald Douglas.
00:01:50
On commercial aircraft for about seven years.
00:01:53
And then kind of realized that, um, sitting in front
00:01:57
of a computer and doing finite element
00:02:00
analysis wasn't exactly something I wanted
00:02:02
to do for the rest of my life.
00:02:04
And so I decided to go to grad school
00:02:08
and get my MBA
00:02:10
and at the time, all of a sudden, high tech
00:02:13
and Silicon Valley was, you know,
00:02:15
of tremendous interest to me.
00:02:18
And, so I ended up aiming really high.
00:02:21
There was a company at the time, in the early nineties -
00:02:24
it's called Silicon Graphics, SJI,
00:02:26
and it's like, 'I wanna work there.'
00:02:28
And, so I applied to SGI
00:02:31
and several other companies.
00:02:32
And, you know, it was a stretch for me.
00:02:35
I mean, I was an aerospace engineer,
00:02:37
and these are companies, this is a company that was making
00:02:40
graphics for making movies.
00:02:41
And so I was fortunate enough to get a position there.
00:02:46
And that was really
00:02:48
what changed the trajectory of my career.
00:02:50
Like, I went from a very conservative aerospace industry
00:02:55
to a very agile, very fast moving high tech.
00:02:59
And so, you know, after several years there
00:03:03
and a bunch of startups,
00:03:05
I was working at a startup as chief marketing officer,
00:03:09
and it was machine learning
00:03:12
and AI for industrial world.
00:03:15
And throughout these years, I stayed in touch
00:03:18
with my boss at SGI, and so he decided to retire
00:03:21
and move back to the UK,
00:03:23
and he recommended me for the position.
00:03:26
And, I knew at the time,
00:03:29
and this is seven years ago, that AI was a future
00:03:33
and was gonna take off.
00:03:34
And if there was one company that was gonna lead
00:03:36
that it was going to be Nvidia.
00:03:38
So I jumped on the opportunity
00:03:40
and was fortunate enough to
00:03:43
get the position that I have today.
00:03:44
Yeah. And, and so in your story, there is a lot
00:03:48
of risk taking, a lot of courageous moves from Spain,
00:03:52
deciding you love physics, love math to aerospace,
00:03:55
and then the next step, as you said, graphics
00:03:57
and then chief marketing officer.
00:03:59
I actually didn't know that I started at the degree in
00:04:01
marketing and have worked my way into business,
00:04:03
and now I'm talking about AI.
00:04:04
So completely coming at it from a different direction.
00:04:07
So maybe as we get started, let's, let's talk about AI,
00:04:10
what it is.
00:04:11
And, and what's interesting is I hear people say, "Well,
00:04:14
I don't think I'm, I'm going to use that,"
00:04:16
or "It's not for me."
00:04:18
or "I haven't learned a lot about it."
00:04:20
but reality is most people already have AI in part
00:04:24
of their life, whether they know it or not.
00:04:27
And it's around us. It's in our phone.
00:04:29
It's in devices at home.
00:04:31
Talk about some of the more practical ways that
00:04:33
you know right now AI is in our lives,
00:04:36
and let's make sure everyone here is using these models,
00:04:40
using these applications is way easier than when we started
00:04:43
30 plus years ago in my career. Computers were hard
00:04:47
and it took a lot of training and a lot
00:04:48
of time on call centers, and it's all gotten better.
00:04:51
So it's actually easier to learn now than it was then.
00:04:53
Yeah, no, absolutely.
00:04:55
I mean, AI is basically training models
00:04:59
that can do certain things.
00:05:01
And a perfect example is if you're
00:05:04
watching TV a lot and you use Paramount plus, it knows
00:05:07
what kind of movies you like,
00:05:09
and it's constantly recommending personalized
00:05:11
recommendations based on your taste and your history.
00:05:15
Or when we're shopping,
00:05:17
if there's a good personalization e-commerce solution out
00:05:20
there, it's actually recomm... knows you, knows
00:05:23
what you like, and it's recommending the right products
00:05:26
that would be of interest to you.
00:05:28
And, it's helping you, um,
00:05:32
pick what you're looking for.
00:05:33
It is, and all the data that we enter,
00:05:36
and the more we tell it, obviously what we try
00:05:39
to do is understand your intention.
00:05:41
So as a shopper, when you express intent,
00:05:43
what you're looking for, you may be asking about services at
00:05:47
the auto care center or ordering a cake at the deli,
00:05:50
or you may be ordering something of a certain color.
00:05:53
We try to help make it more intuitive so
00:05:56
that the next time you're in, we can
00:05:57
know those things about you.
00:05:59
We can know what kind of car you're in.
00:06:00
We can know roughly how many miles
00:06:02
that is suggest it's time for service.
00:06:04
Or if we know that you love the color red
00:06:06
or you're a certain size,
00:06:07
then we bring back some more choices
00:06:09
that hopefully make your shopping experience
00:06:11
easier and more intuitive. Yeah. And
00:06:13
I wanna encourage everyone to use
00:06:17
Chat GPT or Perplexity for simple things like
00:06:22
um, I wanna throw a party for my daughter,
00:06:25
and she loves mermaids.
00:06:28
You know, what, what should I, why should,
00:06:30
what should I be buying
00:06:32
or how do I organize a party
00:06:35
and give you ideas about simple things like that
00:06:39
or writing a Valentine's note
00:06:44
for your significant error.
00:06:45
And you'd be amazed that the great recommendations
00:06:49
that Chat GPT or Perplexity can make.
00:06:53
So I encourage everyone to go ahead and use it,
00:06:55
because it actually can be an incredible
00:06:59
assistant to anyone.
00:07:01
Let's talk about Nvidia for just a second
00:07:03
before we get into the technologies that are out today.
00:07:06
Started in late nineties, if I remember right,
00:07:09
In video architecture for gaming, is
00:07:12
that correct? Yeah,
00:07:14
I, we say it that right mean, basically it was about
00:07:16
making gaming almost like real life, like you're
00:07:20
inside the game, and for that you needed a lot of graphics
00:07:24
power and rendering.
00:07:25
And so Nvidia basically started in the
00:07:30
graphic processing unit for rendering of video games.
00:07:35
And in fact, when I joined six years ago,
00:07:37
people didn't know who we were.
00:07:39
It was like, oh, you're the gaming company.
00:07:41
And we're like, no, we're like the AI company now.
00:07:44
Everybody's like, you are the GPU company.
00:07:46
It's like, no, we're actually the
00:07:48
accelerated computing company.
00:07:49
Then this year, in January,
00:07:52
Nvidia was the keynote at CES.
00:07:54
Jensen was there, and back to rendering
00:07:57
for a second, he showed a video.
00:07:59
And in the video we're creating about one out of every,
00:08:02
I think at one out of every 32 pixels.
00:08:04
So we're, we're creating a bit of it.
00:08:06
And in the machines, the models can run
00:08:08
and create the rest in real time,
00:08:10
which is really different than playing a video game,
00:08:12
which is reacting to inputs.
00:08:13
Right now, it's doing it on its own.
00:08:15
Quite, quite a big journey from back
00:08:17
in the nineties to where we are.
00:08:18
Yeah. Yeah. And the other thing is Agentic AI.
00:08:22
Right? Jensen talked about Agentic AI,
00:08:26
and these are agents
00:08:30
or generative AI models that you train,
00:08:33
they're specialized
00:08:35
a specific area of expertise.
00:08:39
And they kind of work in concert
00:08:41
and orchestration with other models
00:08:43
and agentic AI models, they can perceive.
00:08:47
It's almost like they can see.
00:08:49
They can reason. They think,
00:08:51
and they take tasks, and they plan,
00:08:56
and then they're autonomous.
00:08:58
They take action.
00:08:59
So they do specific things
00:09:01
or they trigger other agents to do specific things.
00:09:05
And I think the example we talked about at NRF
00:09:08
and I think it's simple to understand is a agentic,
00:09:12
AI agent that's watching the weather,
00:09:14
and if there's going to be a snow storm coming in 48 hours,
00:09:20
it's instructing other agents to go and order more shovels.
00:09:23
Or flashlights
00:09:25
or so forth to make sure that you have that in the stores
00:09:29
that are going to get affected by that.
00:09:31
And of course, that thing, agentic AI in itself is,
00:09:35
is really evolving very, very rapidly.
00:09:38
And the way I'd like people to think about it is,
00:09:42
they're creating digital intelligence
00:09:45
and they're bringing knowledge faster
00:09:49
to your associates.
00:09:51
They're helping your associates make decisions much,
00:09:54
much faster, and helping your company become much more agile
00:09:59
and faster response to changes that are happening,
00:10:04
that impact your business and
00:10:07
Wally is helpful because you can just ask a question
00:10:10
and it runs off and it finds all the data,
00:10:12
then it summarizes it
00:10:13
for you in a way that's easier to understand.
00:10:14
So it's faster. It's real.
00:10:17
It gives you more real time data and,
00:10:18
and true insights into what's going on in these models.
00:10:20
We talk about perception
00:10:22
and then reasoning, and then taking action.
00:10:24
Perception is just as you said, seeing, hearing,
00:10:26
gathering what's going on around you.
00:10:28
Reasoning is then thinking through all the choices
00:10:30
that you could make, and then at a point then we can
00:10:33
decide to allow them to take action.
00:10:34
But for retail associates, let's talk about some
00:10:37
of the examples like drawing planograms,
00:10:40
modulars for stores.
00:10:41
That's an example
00:10:43
of we gather the data. We decide
00:10:46
what should be on the counter, on the shelf, how
00:10:48
to best arrange it, and they can draw it out.
00:10:50
And then that allows our associates to be the editor.
00:10:53
In stores it can be
00:10:55
things like we've looked at everything happening in your
00:10:58
store when you walk in the morning,
00:10:59
here are things going on, and here's some things
00:11:01
that could most readily
00:11:03
or readily there to be able
00:11:05
to help you have the best day you can have
00:11:07
and the most productive day you can have.
00:11:08
So just saves a lot of time
00:11:10
and a lot of fact gathering.
00:11:12
And, you know, most people may not know,
00:11:13
but our supercenters are about four and a half acres,
00:11:16
and some of our distribution centers are over 20.
00:11:18
And so having this information when you get there
00:11:22
saves you a lot of time in walking around,
00:11:24
gathering what you need to.
00:11:26
'cause in a lot of cases, the data's there.
00:11:28
We just haven't been able to have in the most simple terms,
00:11:31
the computation to be able to get it and bring it back
00:11:33
and present it in a way that's
00:11:34
in the language that you speak.
00:11:36
It could be English, Spanish,
00:11:36
French, whatever you want it to be.
00:11:38
I mean, I would just say AI is
00:11:40
so much bigger than agentic AI, right?
00:11:43
Like, computer vision has been around for so long,
00:11:46
and, it's being used in stores
00:11:50
for example, for
00:11:52
loss prevention in the back room, making it easier
00:11:55
for customers to check out.
00:11:58
I think the planogram example that you talked about,
00:12:01
that's more like physics AI, right?
00:12:03
Being able to create a digital,
00:12:06
physically accurate digital representation of your stores
00:12:09
and actually integrating your planogram in there
00:12:12
and be able to have your merchandising people actually look
00:12:16
at different planograms
00:12:17
and decide what is the best planogram digitally,
00:12:20
and how does it really sit on your
00:12:24
shelves before they actually
00:12:26
go and make that decision.
00:12:28
And so it optimizes ultimately your merchandising
00:12:31
and your revenue.
00:12:33
And then of course, all the examples
00:12:34
that we talked about around generative
00:12:37
Ai. Yeah. Someone, someone
00:12:38
told me recently that our,
00:12:39
our limitation is our imagination.
00:12:41
So with the evolution of what we're going through,
00:12:44
we're talking about from machine learning to AI,
00:12:47
to generative AI, agentic, physical lots after omniverse.
00:12:51
Like if you can dream it, it can probably happen.
00:12:54
But the, what is important to us is we are,
00:12:57
we're people led and we're tech powered,
00:13:00
and we're an omnichannel retailer.
00:13:01
We have a strong purpose
00:13:02
to save people money and live better.
00:13:03
Our founder gave us that. That won't change.
00:13:05
We have core values that we're really proud of.
00:13:07
Those won't change, but we want to be able to give our
00:13:10
associates all over the company, in stores,
00:13:13
fulfillment centers, distribution centers here in the home
00:13:15
office, the very best tools so that they can start their day
00:13:20
with the information they need,
00:13:21
and they can just move into the day of momentum
00:13:23
and not have to relearn, redo some of the things
00:13:25
that they've had to do in the past.
00:13:26
Because as you said, the more data,
00:13:28
the more practice, the models just get better.
00:13:30
Yeah, absolutely. I mean, we seriously believe that
00:13:35
AI is assistance
00:13:37
To your associates.
00:13:38
It's giving them the information that they need, you know,
00:13:41
in real time so they can make smart decisions,
00:13:45
make recommendations with a lot of data.
00:13:48
And so they can spend their time helping your customers,
00:13:52
you know, find exactly what they're looking for, making sure
00:13:55
that the store is set up in the best way to provide the best
00:13:59
shopping experience for your customers.
00:14:01
So it's really about digital intelligence
00:14:04
and productivity for your associates. That's
00:14:06
Exactly right. Okay.
00:14:07
Last thing.
00:14:08
If you're new to the subject,
00:14:10
or if you, if you don't know anything about generative AI
00:14:14
or you, you haven't heard any of this
00:14:15
before, start with, what would you recommend?
00:14:18
I've met you a few times now, I can tell you are one
00:14:21
of those people that probably learn something every hour
00:14:23
of every day, and always curious,
00:14:25
but it's never too late to decide you want
00:14:29
to get into the field that you're in, do what I'm doing,
00:14:32
but it's never too late to try
00:14:34
to just jump in, learn, pick up and go.
00:14:36
And to me, this is an interesting time
00:14:37
because so much of this is new. People, associates, students,
00:14:42
anyone around the world can make a decision to be a part
00:14:45
of it, and there's so much opportunity for everyone.
00:14:47
Is that the way you feel about it?
00:14:48
Yeah, absolutely. I mean, I think my number one advice
00:14:52
to everyone is, first of all, all this
00:14:55
AI stuff is new, right?
00:14:57
Some of us have been in it longer,
00:14:59
but you can learn about it now
00:15:03
and you can start using it.
00:15:04
And you don't have to be a data scientist to leverage an AI.
00:15:09
There's so many incredible tools out there, you know, Chat GPT,
00:15:12
perplexity,
00:15:14
that can make you actually a lot more productive.
00:15:17
So, my recommendation is
00:15:19
the world is changing at incredible pace.
00:15:22
And, if you have the attitude of,
00:15:27
I'm too old for this,
00:15:29
you're like doing yourself a real disservice.
00:15:31
Like, it doesn't matter what age you're at if you're really
00:15:34
young or if you're much more mature,
00:15:37
there's so much to learn to really leverage that be
00:15:41
an ever-learning machine in my opinion.
00:15:45
And stay abreast of all the latest that's going on.
00:15:50
Listen to your gut. Like what gets you excited
00:15:54
and interested and go pursue that.
00:15:57
And I always tell people, don't
00:16:01
count yourself short.
00:16:03
I would especially say this to younger women.
00:16:07
There are a lot of times when
00:16:09
there's a position open
00:16:11
and you go, well, I meet six of the seven
00:16:14
qualifications, so I'm not gonna apply.
00:16:16
And you know, I encourage
00:16:19
every young person in particular to
00:16:22
believe in themselves, believe in
00:16:25
that you are not always gonna know the answer to everything,
00:16:28
but you're smart enough
00:16:29
that you will learn if you're passionate about it.
00:16:33
And so go for your dreams.
00:16:35
Uh, try to learn as much as you can from other people
00:16:38
that you know could be role models for you.
00:16:42
And, you'll get the job.
00:16:44
It's happened to me multiple times.
00:16:47
And, just as long as you're staying smart
00:16:50
and learning, you will grow in your career.
00:16:52
Don't be afraid of failure.
00:16:54
- Failure is the best way to learn.
- It's temporary.
00:16:57
It's a learning opportunity.
00:16:59
And it's a wonderful learning opportunity.
00:17:01
Yeah. Really, really great advice.
00:17:02
And, and if you apply and you don't get it, it's okay.
00:17:05
Apply for the next one. Exactly.
00:17:06
But, but the best way
00:17:07
to be sure you'll get a no is not to try.
00:17:10
And someone told me one time that
00:17:13
losing or, you know, failure is temporary.
00:17:16
But not trying or giving up that's permanent.
00:17:19
It's what did they say?
00:17:21
It's no now, but maybe later.
00:17:26
Well, thanks for coming. We're going to get
00:17:28
to spend some more time with more associates today.
00:17:31
I'm sure they're gonna love being around you.
00:17:33
But for everyone out there, just know that
00:17:36
we're all thinking about
00:17:37
and working on how we can make the future
00:17:40
of our company more successful with partners like Nvidia
00:17:44
and partners like Azita.
00:17:46
Most importantly, wanna make sure
00:17:48
that we are developing the right tools so
00:17:50
that we can help you be the most effective
00:17:52
that you can be in your role
00:17:54
with all the information we can be.
00:17:55
Thanks for listening
00:17:56
and looking forward to the rest of the day.
00:17:58
Thank you. Thanks for having me.
