Accelerate your business transformation using the power of AI with Microsoft AI Cloud Partner Program. Discover the benefits of Azure and Azure AI, learn about the differences between AI and ML, and explore real-world use cases of cognitive services. Start your free account and digitally transform your business today.
The video discusses the benefits of Azure and Azure AI, highlighting that Azure is a secure and trusted cloud platform for data. The video also emphasizes the role of Azure AI in transforming businesses digitally. The conversation between Tom O'Reilly and Ali Dalloul focuses on the differences and similarities between artificial intelligence (AI) and machine learning (ML). They explain that AI is a broader term encompassing various cognitive services that mimic human behavior, such as computer vision, speech, language, and decision making. On the other hand, ML is a subset of AI that focuses on areas where cognitive services are not applicable, enabling enterprises to build their own AI models and leverage machine learning for tasks like fraud detection. The video also highlights real-world use cases of cognitive services, such as Netflix's personalized content recommendations and Tesla's AI-powered autonomous driving software.
- When we talk about Azure plus Azure AI,
we're saying we have a super computer cloud.
It is the best place for your data
that's also trusted at scale.
And when you take Azure AI and Azure together,
we are here to help you digitally transform your business
into the AIH.
- [Announcer] Welcome to the
Microsoft Cloud Executive Enablement series,
where we speak with Microsoft Cloud senior leaders
and experts about the latest trends in technologies
we are seeing in the market.
The goal of this series is to share with you and your teams
our perspective on the business value driven
by the Microsoft Cloud for our mutual customers,
and the opportunities for our partners to grow
their business with Microsoft.
- Hi, I'm Tom O'Reilly.
I work with the Microsoft Partner Development team
looking after data and AI.
And today's episode, we're going to be talking about AI and ML.
And to help me with that conversation, I have Ali Dalloul
who's an expert in this topic
and has been around Microsoft for 25 years.
So, Ali, firstly,
thank you so much for joining us, and welcome.
- Thank you, Tom.
- Let's start with some definitions, right?
Whenever we see AI, we often see a /ML written
straight after it.
What's the difference when we're thinking
about artificial intelligence versus machine learning?
You know, how are they different?
How are they similar?
Why are they always put together?
- Yeah, great question, Tom.
And they're actually not that dissimilar.
So artificial intelligence is really the umbrella term,
and machine learning is a subset of that.
So let's unpack that a little bit.
So, artificial intelligence really truly means
that the way software can mimic human behavior
evolves in a way where it learns.
We look at the physical world, so we see,
that's computer vision.
We speak, that speech and language.
We translate if we have the ability to translate,
and we know multiple languages.
That's machine translation.
We hear and we comprehend,
and that's machine reading comprehension.
So all of these,
we call them the cognitive services side of AI.
So these are the kind of artificial intelligence
where we pre-trained the model, right?
And they're basically based on key pillars
of vision, speech, language, and decision making.
And we'll talk about OpenAI,
which is kind of a new category
of a Azure cognitive services,
which is really what we call general purpose AI.
So these cognitive abilities are pre-trained.
Machine learning on the other hand,
it is a subset of AI,
and it's just more focused on categories
where cognitive services or cognitive abilities
are not applicable.
It's really more DIY
and where you really want to be able to do AI at scale
in areas that don't have the natural patterns
of human behavior.
For example, for credit card fraud detection, right?
It's not necessarily a human cognitive ability.
It's not about seeing or talking
or speaking or comprehending.
It's about determining patterns at scale
on large volumes of data,
but learning from the mistakes.
The old software doesn't learn.
So hence, the machine is now learning.
Over time, it is improving.
The model over time becomes a great model.
So it can detect credit card fraud at scale.
It can help you address network anomaly detections.
It can help you address defects.
British Petroleum is using our machine learning
to look at faults in their reservoirs.
We have Shell using it to determine faults in their sensors
across all of their manufacturing.
So the machine is learning these models.
So enterprises, we provide the tooling
that they can go and do their DIY AI.
They can go and build whatever AI they want
with the right tooling
because we give them that capability.
So machine learning in summary, is really part of AI.
It is a subset of AI versus cognitive services
or cognitive abilities of AI that mimic,
truly human behavior.
- So we were talking a little bit
about the cognitive services and those parts
that are related to human capability,
whether that be speech or vision or so on.
What are some good use cases?
And if you can contrast in the industry before and after,
using some of those cognitive services,
what does that look like?
- That's a great question, Tom.
So let's take a look at some of the very, very near
and dear consumer services that you and I
and millions of people around the world
have experienced, right?
Let's take a simple one, okay, the film industry.
Kodak, digital camera, and so on.
So AI today, when you look at Netflix
and their ability to personalize content,
that's all machine learning.
That started with their DVD business
when they started mailing the DVDs,
but they soon realized,
credit to Reed Hastings and his team,
they soon realized that if they don't disrupt themselves,
they're going to be disrupted.
They moved immediately to streaming,
way before anybody believed streaming
is going to be the way forward.
And they were able to establish a beachhead business.
They captured the lion's share.
But a lot of that actually today is, behind the scenes,
is actual AI.
Of course, content is king, but it's AI
that's really truly providing the recommendations,
the personalization, and they're taking that
and they're learning from that
and they're building even more content
because they're seeing the patterns
of what are the consumers using.
Why is Tesla worth more than the entire automotive industry
in the US and maybe the world?
One company. Why?
Because fundamentally, Tesla was built
as a software company.
What do you do when you buy a Tesla?
You sign up to a computer on wheels
that captures every bit of data as you're driving that car.
It's got cameras everywhere.
It's mapping the streets.
That mapping of the streets is what's feeding
into the autonomous driving software,
which is, by the way,
the most profitable module you buy on top of the car
because it's software.
Software has very high margins.
Once you build that IP, your cost is sunk.
But that software is 100% AI
because it's feeding in from all the cameras on a Tesla.
It's picking up from the world around you.
It's learning. It's improving.
It's the only company today that has cars
that have driven billions and billions
and billions of miles, mapping every street
and every permutation.
A pothole in the street,
a tree here, a bridge fallen here, a closed road,
the storm, and so on.
And now, over time, that is a differentiation
and an advantage that that's where
the automotive industry is behind.
That's what people don't understand,
why is this such a success story?
It's because of things like that.
Your audio and your conversation in the car.
You are continuously talking to an AI model.
When you're talking even to your phone,
that is a continuous speech model
that's being trained and learned.
So these companies who are able to take the data
in a compliant way, that data, apply AI algorithms,
they're able to disrupt very well-established industries
in automotive, in transportation.
So these are a lot of different industries, Tom,
and where AI is like, people don't understand
that the core was AI.
Like that disruption was, at its essence,
the business model, the platform
is AI-driven along with cloud.
So that's why when we talk about Azure plus Azure AI,
we're saying we have a supercomputer cloud.
It is the best place for your data.
It's secure, it's compliant.
Your data is your data.
Absolutely, your data is your data.
We don't tap into your data.
We don't use your data to train our models,
but we give you AI that's also trusted at scale.
And when you take Azure AI and Azure together,
we are here to help you digitally transform your business
into the AI age.
- When you're meeting a customer,
they're sort of toying with the idea
of maybe I should think about using AI,
or where will AI provide me the most value in my business?
What's the two or three questions that you ask
that business leader?
- Great question.
So it starts with, it depends who we're talking to, right?
If we're talking to a truly a business leader, as you-
- Let's say the CEO of the company.
- The CEO, the first thing I ask is,
"Tell me what keeps you awake at night?"
And depending on their answer,
I will be very honest and transparent,
as we always should be,
whether AI can solve that problem or not.
Some cases, AI cannot solve it.
But let's say that that business problem
is customer service.
100%, AI can make it better.
Let's say that problem is process improvement,
fraud detection, workforce productivity,
a lot of these different scenarios, okay?
It absolutely can be addressed with AI.
Look at even financial services with robo-advisors.
What do you think a robo-advisor is?
It's an AI bot, right?
That basically takes in all the financial data,
and machine learning applies an algorithm
of what a typical business financial analyst does,
and says, here's a basket of stocks,
or here's a set of funds or index funds
or whatever that match your personality, your risk profile,
the amount of money you want to invest.
If I am a financial advisor as an example, right?
And a robo-advisor can help do 60,
70% of my work, that's amazing
because then I really focus where my core value proposition
to my clientele is going to be,
which is the human relationship,
the trust, the depth of understanding of the law,
of the regulations, of the policies, of the changes,
and really working through with that client
and that relationship.
And AI is just another tool.
The same way I've had Excel before and a spreadsheet,
I have yet another tool.
So we look at these aspects.
Another aspect, for example,
let's take if we're talking to a CIO, right?
So actually, I was talking to a CIO in Asia yesterday,
but one of the things he asked me,
he's like, "I'm worried about my developer productivity.
What do I do?"
I said, "Have you seen Copilot from Microsoft?"
And he's like, "No, what, what is that?"
I said, "Let's show you."
And you know, Copilot is based
on our latest general purpose AI
that's built on OpenAI models and algorithms.
And it does produce code from natural language, right?
So you just say, "Write me a program
that plays tic-tac-toe."
That's it. That's all you have to say.
Plain English, natural language,
and it produces all the code, right?
So when we look at a developer, right?
And if you've ever developed software,
a lot of the code is repetitive.
Why would you want to waste your time
on so much available repetitive code
that you could have actually used from somewhere else?
And you just apply your expertise, your fine tuning,
your optimization, you're securing that code,
putting it into the right context,
what we give you, that's out of the box and AI.
And that gives you the ability to really accelerate
the productivity of the developer.
So when, depending on the persona that we're talking
to, a CIO, a CEO, they all have different business.
So we start with that.
What's the business problem you're trying to solve?
What keeps you awake at night?
Is it something you're willing to invest in
and put in the money to make sure
that you're getting a return?
And we will walk them through.
Here's a scenario would look like in terms of ROI using AI,
assuming again, what that AI
can actually solve that problem.
And once we walk them through that, we go through
the typical cycle of great, I love it.
Let's do some envisioning.
Let's do some ideation.
Let's look at these four or five use cases.
Now let's start getting together and doing a pilot,
proof of concept.
Like, show me it can really work.
I believe you, but show me it can really work, right?
And here, there's an opportunity also where partners come in
because that's where these POCs,
these pilots require tremendous technical capability, Tom.
So we want help in that last mile,
and we want people to come in and help us.
- Outside of the call center,
the fraud detection, the anomaly work,
what are some of the other most common use cases
that you're seeing now?
And what do you see that changing
in the next couple of years?
How do you see that evolving?
- Before we go there, Tom, it's a great question.
I kind of want to answer your question backwards
because it's important, right?
I think today when we spoke earlier about cognitive services
and what we call mimicking human abilities
through pre-trained models.
And we call that narrow AI.
That is a task-specific traditional AI.
It's a bounded problem.
That's why when you're talking to your car or your phone,
you cannot just talk to your car
about what should I have for dinner tonight?
It doesn't understand.
Now, of course, it's using AI,
it's using speech and so on, but it's a bounded problem.
This continues to be mainstream, right?
And that's what we call traditional AI, okay?
Task-specific, narrow use cases.
And some people, you'll hear the word weak AI.
They call this weak AI
because you need a lot of data upfront.
You got to label to train it.
And in that, you really are looking at use cases
around speech recognition and speech transcription.
You're looking at digitalization of documents.
Taking a physical document, applying computer vision,
optical character recognition, applying layout extraction.
We process billions of documents per month,
just doing stuff like that for a lot of enterprises.
And it's very, very high value.
Microsoft itself, you know, Amy Hood,
our CFO challenged us two years ago.
Can you run your forms recognition AI
to help me reduce my costs for invoice automation
and employee expense reporting?
A member of my team won, actually, Amy Hood's award,
when we proved that we saved her 10,000 man hours per year.
So today, when you submit your expense report,
you see it's all automated.
If you've been at Microsoft for a long time,
you know how it used to be.
And now, it's all automated and we've saved
tens of millions of dollars for the company
and 10,000 man hours per year.
So these are like critical use cases
and business process optimization.
If you're using Teams, if you turn on transcription,
that is coming from our AI.
By the way, if you type in today inside Teams in the chat,
and you type in a different language,
like you type in in Chinese
or you type in French or German,
you can right-click that message
and we can translate it in there.
So these are examples
where you're like, we're driving massive productivity
through AI as a productivity company.
These are things that you'd have paid enormous amount
of money for a certified translator.
You got to go through a process and so on and so forth.
So there are a lot of different areas
where you see cognitive services being applied
in our first party product,
in our own Microsoft products across the board.
And you see AI being applied also in different enterprises.
I mentioned examples like British Petroleum.
We're working with also examples like the NHS in the UK,
where they're looking at medical records.
Working with Ernest and Young,
where they look at tax forms and they apply digitization,
optical character recognition,
OCR and layout extraction, and so on.
A lot of companies around the world
are benefiting from these use cases
where you really are improving a business process.
You really are improving, for example,
you mentioned the contact center as an example.
That is one of the biggest areas.
So and Allstate came to us and said,
in a situation where there's an accident,
people are very distressed.
They're very emotional, they're breaking down, anyway.
Hopefully there's no bodily injury.
But nonetheless, even if it's a fender bender,
people are just, there's a shock, there's a distress.
They call the call center at Allstate,
and sometimes, they're mumbling,
sometimes they're not clear.
That agent on the other end of the phone
is trying to understand,
okay, tell me exactly what happened.
They're not hearing them well.
So then there's more frustration.
"Tell me again, I missed that part."
And so on.
So put in Azure AI speech recognition
and speech transcription.
So we recognize, in a compliant way with an opt-in.
What that customer in distress is saying.
We transcribe that.
That service agent doesn't have to take notes, right?
It's immediately transcripted in front of them.
They understand everything.
They don't have to repeat.
That call is shortened.
And immediately from that workflow,
they're applying it to address claims
and make sure that that person's going home safe
and that their claim is being processed pretty quickly.
And they saved, in year one, over $14 million,
just on something that you think is simple.
But that is intelligent automation at scale.
These are real scenarios.
We do it also with a lot of the key ISVs
in the contact center space and many, many others.
So these are real things.
Now, the other part of your question
is so where is it going, right?
So this traditional AI continues, it's very important.
It's core, right?
And it's durable and it solves real problems.
The buzzword these days, or the excitement
these days is what?
You hear is OpenAI.
And Azure OpenAI is what we launched last week.
So let's talk about that if you're okay, Tom.
Why is OpenAI such a hot thing,
and why did Microsoft and OpenAI come together
to be able to build the most intelligent AI out there?
Hype aside, excitement aside,
let's look at in what is general purpose AI.
General purpose AI, unlike traditional AI,
doesn't require the data per model upfront.
It takes the web scale data,
so all of the data of the world, and then some,
and applies massive computing power.
That's why all of the OpenAI models
are actually built on Azure.
Azure powers all of the OpenAI models
and then applies deep science
and the breakthroughs in deep learning and algorithms,
which is kind of really where the core
of the OpenAI models are.
And that allows it to be able to provide a set of models
in two categories.
The language category.
And these are the things
we may have heard called GPT, right?
So these are general pre-trained transformer models.
This is where you also heard the term generative AI.
And where AI is generating, AI is reaching human creativity.
Because now we're moving towards a general purpose concept,
where you're taking all of this data available
and we produce GPT models
that are able to provide language abilities,
where naturally, anything you and I can talk about,
literally anything we, like the way we're interacting
right now as humans naturally, unbounded, unbounded, right?
And that's a very important concept
because as we said in your car, it's bounded, right?
Open the window, turn on the radio.
Take me to.
You can't say, "Get me a pizza."
But now, it's unbounded,
but these are now what we call strong AI.
And now they're more horizontal in nature.
So some of the use cases we were talking
about before were narrow, vertical in nature,
Now you're horizontal, you are broad.
You're beginning to mimic human interaction.
Hence the magical moment of wow,
I can talk to this thing
and it can tell me about like, I have customers
who are in India and they're like,
"Oh, I asked it about Indian culture and it answered.
How did it know?"
It actually is like it knows
because all that information, is available on the internet.
But before in traditional AI,
you would have to have taken
an Indian-specific cultural corpus of data,
train a model to address a very bounded set of questions
in that domain.
So the old customer service of a chatbot
where you asked the question,
"What is my credit card balance?"
And at some point, by the third question, it's like,
"Sorry, I can't help you.
Let me call a representative."
Well now because it's unbounded,
and you can bring in not just the enterprise data
that is relevant to that experience,
but you have the entire corpus behind it.
And you have these algorithms,
and you have this conversational ability.
It's going to be transformative
across the entire customer service spectrum.
So these models are just amazing, right?
We spoke about Copilot earlier.
That's the next set of models.
This is what called Codex models.
They can translate that natural language into actual code.
So the future, and why Microsoft and Azure AI
and Azure OpenAI came together
is really the opportunity of Azure the cloud,
a very transformative AI experience with OpenAI
now available on Azure,
meaning you have responsible AI compliance.
You have security, you have enroll-based access control,
you have scalability.
Your data is protected.
You have now all of the enterprise grade premium features,
and the SLAs and the scalability of Azure.
And that's why we provide through Azure OpenAI,
all of the models of OpenAI,
but in an enterprise grade scenario.
And that is now unlocking for us
a lot of new set of use cases.
That is just beginning to happen as we speak, Tom, today.
- Well, you've given me the perfect introduction
because you've mentioned two legs
of what I think is the three-legged stool.
And that third leg of the stool being partners.
We've got the power of Microsoft, the Microsoft cloud,
the Azure AI now being OpenAI with that.
None of that lands in customers without the partners
that exist in the Microsoft ecosystem.
Where's the opportunity for them?
You spoke a little bit about governance,
management of these models, the security side,
everything that goes into what we call
a traditional enterprise IT deployment.
Where's the innovation for them?
Where's the opportunity for them to continue to build
on top of these building blocks that we provide?
- Excellent question, Tom.
First, let me say that the promise of Microsoft
to the partners and to the enterprises
is really based on trust.
Why do enterprises do business with us?
Because they trust us.
Why do enterprises put their data on Azure?
Because it is their data.
We believe privacy is a human right, as Satya says.
And your data is your data.
Unlike other companies where that data
can be used and monetized, we do not access your data.
We do not monetize your data.
We do not use your data without your permission.
That promise of trust is fundamental to how we do business
with the enterprises and how we also build software.
So let talk about that second part of how we build AI,
and then I'm going to bring in the partners in that equation.
From day one, we understood the power of AI,
and we understood the potential opportunities,
but also the potential risks involved.
And we focused on, we were the first company
to basically talk about responsible AI.
And we set up a framework of AI fairness,
AI accountability, AI transparency, AI privacy,
AI security, AI inclusiveness.
All of that wasn't just policy talk.
We proceeded to setting up the Office of Responsible AI
under our president and vice chairman Brad Smith.
That is a office that actually has a lot of power.
They actually review our work.
They review what we call sensitive use cases.
They can actually block a business
even worth hundreds of millions of dollars
if they say this is going to cause reputational harm
to Microsoft or to the partners
or to the customer.
I'll give you one example.
We had a bank that came to us
and wanted to use speech recognition
in their contact center,
but they wanted to use it in a way
where it is what we call a dragnet,
meaning there's no opt-in.
They don't tell you,
but your voice is your password.
And the way they capture that is they capture everything
you're saying and then they determine
that what you said is really Tom or really Ali.
Now nothing is wrong with that.
That's actually a technology we have.
What is wrong with it is it's not prompted,
it's not opted in, and it could be a dragnet.
Now imagine in that conversation, my 13-year-old
or my 12-year-old is standing next to me and talking.
Now you just now violated federal law
because now you got a minor in the conversation, right?
So we went and educated them,
and we actually saved that bank millions of dollars
of liability and lawsuit.
And they did it the way we told them.
So we have an Office of Responsible AI
that works with us very closely.
Then we went a step further.
We established an ethics committee.
It's called Aether.
AI ethics and effects in engineering.
And they review also all of the work we do.
Then we want a step further.
And we told customers and partners,
we want you to keep us honest.
We want you to keep us challenged.
And we published a set of tools.
So one of the challenges also in AI
is model explainability and interpretability.
What happens to this model when it's out there?
There's areas also of bias.
For example, in a loan application,
in a robo-advisor application,
there could be bias.
In a computer vision model,
there could be bias based on race.
There could be indirect profiling or things of that sort.
So we've applied, we have a consortium,
and we have tools like Fairlearn.
We have InterpretML, which we publish on GitHub and so on.
So then we applied the, yeah, one more thing,
which is called RAISE,
which is Responsible AI and Systems Engineering.
So we build in these responsible practices down to the code.
So we engineer from the grounds up with trust.
We engineer from the grounds up
with responsible AI practices.
We engineer, we further protect with policy.
We work with internal experts and legal and external.
And that is also part of the value proposition
of why Azure AI and why Azure.
So let's talk about the partners.
Now, the partners are an extension of that, Tom.
Why? On three levels.
The first is with all the goodness and excitement
and my passion about AI, it's still complex.
You're not buying a phone or Windows Surface.
You're buying very complex technology.
And this complex technology in the enterprises,
just like other technologies, requires expertise,
requires a deployment techniques,
requires fine tuning techniques.
Here's the opportunity and the challenge.
The challenge is because it's complex,
it requires specialized talent,
which is why AI people are in high demand, okay?
Data scientists, engineers, product people.
They're in very high demand.
Business people are very high in demand.
Partners who have that expertise in-house
have an advantage in the market.
Because even our largest customers have a limited set
of that talent in-house.
And even if they have, they're competing
with the likes of Microsoft
and others for that talent
because it's in very, very high demand
and very short supply.
So partners who have the talent, who have the relationships,
who have the trust, who have the know-how,
and can do system integration
and really understand the business sales cycle,
they're in an advantageous position.
Coupled with our enterprise offerings,
coupled with what we are providing from a trust perspective,
they are the last mile.
They're the most important mile.
Imagine now all this AI, it's sitting on the cloud,
but I cannot get it in the hands of my enterprises.
Well, I have no distribution.
I mean, they are the most critical link, right?
And not only that, we are sitting on a massive opportunity
that is, we haven't even yet scratched the surface.
I mean, AI, if you look at different numbers,
it's between 50 to a $100 billion
in addressable market this year alone, right?
And exponentially growing.
So it's a very, very large market.
And this is not even including the hardware, right?
This is just the software.
This is not even including the services.
The professional service is even a larger pie.
So there's an opportunity for these GSIs,
to work very closely with us,
an opportunity for us to work very closely with them
to deliver that value to the customers.
Because A, they have the talent.
B, they have the relationships
and the know how and the understanding.
C, AI pulls through a lot of other businesses.
It pulls in Azure Cloud.
It pulls in our data business, SQL analytics business,
and many, many other businesses.
So third, it is the hottest topic in the C-suite.
So if you really, like a lot of customers come to me,
like even Microsoft Field,
I get so many requests like the CIO or the CEO
just wants to talk about AI.
But there might be like a very large office deal,
but no, no, no, no.
They want to talk about AI first.
Okay, so it's a door opener.
But you got to come in with the right set of messages
and you got to come in with the right set of humility
and grounding, as we said earlier.
Like, that's my opening question.
What is your business problem?
What keeps you awake at night?
And if the partners ask the same question,
and they can match it with their professional expertise
and with our technology,
they can really unlock a lot of value.
I think working with people like you, Tom,
and your team and the partner field organization,
we can generate leads for them.
We can give them a lot of training.
We can give them a lot of enablement.
We can support them in sales calls.
We can support them in marketing events.
We can really help them deliver that last mile value
to the customers.
And we'd love to work with them.
And we cannot do it alone.
Even the size of Microsoft,
Microsoft has, from the grounds up,
been a channel-driven company from the day it was created.
It was, from the day it was founded by Bill,
it's always been about the channel.
The channel is the lifeblood of Microsoft,
will continue being the lifeblood of Microsoft.
And it's very critical for us today, more than ever,
especially with the channel and the SI
that actually have the AI talent.
We would love to work with them to go and unlock that value.
- So that's a great place for us to leave off.
I've loved having this conversation
with you. - Thank you.
- I've learned a lot already.
Final thing for you in the last sort of 30 seconds,
what do you love the most about your job?
- Oh, that's a great question.
It doesn't show so far? (laughs)
- Well, I can see your passion about the topic already,
but what's the thing that-
- Yeah, three things.
It's actually a very deep question.
The first is, I really genuinely am driven
by the power of technology to improve human life.
And anything, not just like let's go solve
this business problem.
I genuinely, it's my sense of purpose.
It's like I wake up every day and I'm like,
if there are no challenges, it's a boring day.
The second is I'm so blessed to work
with the most amazing people in the world.
People like you in the field,
people like our scientists.
How much you learn from these people,
and how much you really can take that and continue to evolve
as a human being.
And guess what?
AI is about evolving and learning.
It's what machine learning, the machine is learning.
The machine is evolving.
So we are also similarly evolving.
And I love to keep taking on the data
and learning myself and evolving.
And the third really is, I mean, how can you not be excited
when things like OpenAI and ChatGPT,
and I mean, the moment we're in right now?
I mean, this is the moment.
This is where we rise to the occasion, both as a company.
If you listen to Satya's earnings call yesterday.
And he called it, this is the new era of AI.
This is the moment, this is the time where AI
in the technology adoption lifecycle
is beginning to cross the chasm.
So a lot of companies die before they cross the chasm.
A lot of technologies die before they cross the chasm.
But when you cross the chasm, in Geoffrey Moore's framework,
you are beginning to hit the mainstream.
Once you hit the mainstream,
which I believe we're on the cusp of right now,
we are just at the right moment,
you'll unlock massive innovation and value.
I'll give you a couple of examples
where we believe with general purpose AI
and with the democratization of AI,
we are at a moment today where models as a platform
is akin to the cloud computing revolution.
So the same example I gave you before,
which is anybody today with a swipe of a credit card
can have access to supercomputing
that the largest governments in the world
did not have before.
Today, you have that with AI.
If you look at companies like Jasper.AI
are worth now $2 billion.
It's a startup that's 100% built on OpenAI models.
We are democratizing AI with access to models
as a platform on Azure
for new grassroots innovation on companies
and startups to flourish and get built.
And for enterprises to capture also that value
and transform their customer experience,
transform how they build products,
transform how they empower their employees,
and really kind of accelerate their journey.
So if that doesn't excite me,
like, I don't know what is going to excite me.
And we're in a time where this macro environment,
as difficult as it is, I'm an optimist at heart,
if you can't tell already,
but AI can help you do a lot more with less.
Because again, it is that intelligent automation.
It's like, we are at a point
where we have these amazing tools.
If I can generate 80% of my code through Copilot,
the thing I'm pitching right now
is like something is going to clean my inbox.
And if I can get somebody to clean my email inbox,
an AI bot that's just going to,
like my productivity is going to go through the roof.
So we are at these times where,
look at what we're doing in Office,
what we're doing in Teams.
I mean, that's the optimism I have for the future.
It's very bright.
My kids started on ChatGPT the day it was out.
They came to me, it was like, I said, "You can't cheat."
You know, like, "No, no, no, no, no.
It's fascinating, Dad."
And so on.
But even education right now.
Even education, they're rethinking education
because of the innovations in AI.
So you and I probably grew up in an era
where even calculators were forbidden in the classroom.
Then we got calculators,
and they were like, oh, calculators are okay.
Then we got computers and now computers are okay.
Now you have AI.
So educators, students have to rethink
what is the new paradigm of learning.
For example, if ChatGPT
can provide this massive conversational ability,
what is one of the most desired skills today
in any organization or in the market?
The ability to communicate,
the ability to stand and talk to other humans.
Well, guess what?
The same equivalent example of something can generate code.
If something can generate the conversation for me,
if it can do a lot of that generation,
I can take that, I can communicate it.
I can become a better presenter.
I can engage in different ways of humans.
That then moves up through the value chain
to do greater things.
So these are the three things that excite me.
Well, thank you so much for joining us today.
It has been an amazing privilege for me to be able to talk
to Ali today about where AI is at.
And this is the moment that AI is beginning
to cross the chasm.
We've had some amazing announcements
come out from OpenAI,
and it's combination with Azure AI
makes an amazing opportunity for our partner ecosystem.
We have opportunities that are coming up every day
from using cognitive services
that we heard about that human ability
to do things like speech and vision,
to a lot of this anomaly detection,
a lot of patent recognition,
which computers are just fundamentally better at,
and the use cases for that that exist
across financial services, healthcare, customer service,
all sorts of different things make it really exciting.
But most of all, we're at a part where we get to reimagine
what this future is going to look like.
And we are delighted to be able to share
and reimagine that future together with you.
Thank you so much for joining us today
and watching this episode together with us,
and I look forward to having you join us
on our next episode.
- [Announcer] And that wraps up today's episode.
Don't forget that this episode is a part of a series
featuring some of our most experienced
and innovative global executives,
packed full of great insights and examples
of how to make the most out of working alongside Microsoft.
If you haven't already,
make sure to check out our other episodes.
No matter your industry or area of focus,
the Microsoft Global Partner Enablement team
is here to enable you and your teams to achieve more.
If you want to hear a little more of this episode,
we have a podcast which has some more of our discussion
and some bonus content.
If there's an area of cloud innovation
you'd like to hear more about,
please send us a note at salesenablement-GSI@microsoft.com
so that we can create content
that meets your enablement needs.
Thank you for listening.
Thank you for engaging with us,
and thank you for being a Microsoft partner.
We'll see you on the next episode
of the "Microsoft Cloud Executive Enablement Series".
Artificial intelligence (AI) and machine learning (ML) are often seen together, but they have distinct differences. AI is the umbrella term for software that mimics human behavior and learns from it. This includes computer vision, speech and language processing, machine translation, and machine reading comprehension. These are referred to as cognitive services, and the models are pre-trained. On the other hand, machine learning is a subset of AI that focuses on areas where cognitive services are not applicable. It involves building models that can learn from large volumes of data and improve over time.
Cognitive services have revolutionized various industries by leveraging human-like capabilities such as speech, vision, and language processing. One industry that has greatly benefited from cognitive services is the film industry or streaming services like Netflix. Personalizing content recommendations is made possible through machine learning. By analyzing user behavior and patterns, platforms like Netflix can provide highly targeted content suggestions, improving user experience and engagement. Tesla is another prime example where AI has transformed the automotive industry. With its autonomous driving software, Tesla captures and analyzes vast amounts of data from its vehicles' cameras, enabling continuous improvement and development of autonomous driving capabilities.
AI also has a significant impact on customer support. Companies can utilize AI-powered chatbots to handle customer inquiries, improving response times and efficiency. These chatbots are trained to understand natural language and provide relevant and accurate answers. Additionally, AI can help detect fraudulent activities, network anomalies, and manufacturing defects by analyzing large volumes of data and learning from patterns. This technology is especially beneficial to industries such as finance and manufacturing.
In conclusion, AI and machine learning are transforming businesses across various industries. From personalized content recommendations to autonomous driving, AI is driving innovation and improving customer experiences. By harnessing the power of cognitive services and machine learning, businesses can unlock new opportunities and stay ahead in the digital age.
Harness the power of AI in marketing for improved efficiency and creative potential. AI tools automate tasks, personalize content, and generate original e-commerce content. The future of business includes shifts to platforms, experiences over transactions, and data-driven insights.
AI offers immense potential for businesses across industries. It can optimize processes, provide personalized services, create intelligent products, and improve efficiency.
AI is revolutionizing customer service with automated chatbots, personalized recommendations, and targeted marketing efforts. It improves efficiency and reduces costs for businesses while providing a convenient and personalized experience for customers.
We appreciate your recent sign up for a LiveAgent.
A message will be sent to your email address containing login details, right after your account is installed.
We’re available on multiple dates