Episode 4: IT IQ Series.

click to begin

Youtube

- [Woman] So Mark, there's so much out there in the world of technology now,
and all these new terms have really come to light in terms of machine learning,
artificial intelligence, bots. What are they? Are they the same?
Do they work together? Can you shed some light on that for us?
- [Mark] Sure. So yeah, very good point. There's a lot of confusion,
particularly around artificial intelligence, machine learning, and
to some extent, bots. The thing they have in common,
which is why people often confuse them, is they all run from the Cloud.
And what the Cloud is about is a massive amount of computing power that's
behind them. But they are very different things in their own right.
So the simplest way to think about it, machine learning is using the power of the
Cloud to come up with answers to questions. So that's really how to think
about it. So machine learning uses data to understand answers to questions that
we want. So if you want to ask a question of some data, you would use machine
learning to do that. Artificial intelligence,
on the other hand, is really looking at the things that human beings might do.
So I'll give you an example. We have a set of services in Azure called
Cognitive Services. And those services can both learn and understand the kind
of things that human beings do. So right now, I would look at you,
at your face, for example, and say, "Are you happy or sad?"
And as a human being, I can understand, if you're smiling, you are happy.
If you're frowning, you might be angry. So AI is using some of those
human-type characteristics, but using the power of the Cloud
to do them. So an example, one of our cognitive services is
an emotion service that could look at an image or a video of somebody's face and
determine whether they're happy or sad, and then you can use that to respond.
And the last thing, bots are really…they're automation.
So when you think bots, it's short for robot, and that's really what they are.
They're robots. They're things that are designed to perform a specific set
of tasks that might not necessarily be repetitive, but is predictable and can
be automated. And the typical use of a bot might be to answer questions.
So you've got a list of things that people might ring up and talk to someone
on the phone. They would ask a question, and the person on the phone would look
up the answer and tell them. A bot could do that.
It could look up the answer and tell them. But all of them use the Cloud,
which is why they're often put together and sometimes confused.
- And now, you mentioned that bots are quite similar in terms of a robot.
But when I think robot, I think something a bit more tangible,
a bit more hardware-based. Is that necessarily the case for bots?
- Yeah. So bots, I guess it's a shortening of the word because a bot is kind of a
smaller robot. A robot is something, yeah, as you said, that might be physical,
that might perform a physical task. Whereas a bot is, yeah, software-based,
based in the Cloud. And most of us have probably used one and not known
it already. So if you've ever been to a website and it's popped up a little window
to say, "Can I help you," and you've asked it a question and
it says, "My name is John," or, "My name is Allison," or whatever,
it probably isn't a real person at all. It's a bot. And when you ask it
a question, it will try and interpret the question and go and find you an answer.
So yeah, it's a software version of a robot. So yeah, very similar.
- So you mentioned around cognitive services that Microsoft does have
a fair offering in that space.
Can you share with us what that is and what some of the use cases might be
in an education instance for such types of technology?
- Sure. So cognitive services, we've got more than two dozen
at the moment. And some examples of those are things like emotion recognition,
things like facial recognition, those sort of things, interpreting words,
so reading people's words and understanding what people are saying.
And one example might be, say monitoring a schoolyard.
So you're looking at a playground. You might get a video feed from a camera,
and that cognitive service could look at the emotions and the behavior in the
playground and determine, for example, if a group of kids might be getting too
aggressive, or they might be angry. They might have raised their voices,
or they're behaving in a way that could be leading to a problem. And in that way,
you could flag that as a problem, perhaps alert somebody.
We could get some teachers there or, you know, at least record what's going
on and monitor whether or not that's a problem. Another way you might use that is
in a distance education environment where perhaps a teacher is giving a lesson
or a lecturer is speaking to a group of students, and those students are working
via video from a distance. And the person giving the lesson wants
to understand, "How are the students reacting? Are they bored?
Are they focused? Are they distracted? Are they angry? Are they happy?"
And get that real-time feedback as to how the lesson is going.
And a last example is language, so the ability to listen
to somebody speaking, turn that into text words that can then be
used a whole variety of ways. Later on, perhaps, someone could search through a
video for a particular term and go back to that point in the video where someone
was speaking, translate it into a different language, or even read it
out loud. So somebody who has a sight impairment can hear the lecturer even if
they can't actually see the person or the presentation.
- Well, that sounds like it's a really advanced technology,
something that certainly has a high level of functionality and capability,
potentially a very high impact. Do you see that here in Australia
there is a certain level of uptake? And is that uptake evenly spread
across our different segments, whether they be primary, secondary,
or higher education?
- Sure. So the answer to the last part first, absolutely. So in fact,
we've been really quite surprised at the extent to which the whole education
industry from primary all the way through to tertiary education have really jumped
at this. And there are really two distinct areas where we're seeing uptake.
One is in the area of STEM learning itself, so actually using these
technologies with students of all ages, whether they're kids in primary school
building robots from robot kits, doing some basic but exciting things
like that, all the way through to university lecturers and researchers using
things like Cognitive Services and AI to really power some dynamic outcomes
in research. So there's that whole area of what the students are doing themselves and
how we're using those technologies to enhance teaching and teach kids
skills that, quite frankly, in the future, they're going to need.
So there's that area, and then there's the business
of education itself. So as the example I gave earlier,
how can we use some of those technologies to actually reduce cost or,
more importantly, to improve the way that we run our schools and universities?
And it can be anything from monitoring power and understanding better ways
to save money around power and electricity or energy, right through to, as I said,
it might be bullying behavior that we're looking at. It might be using those
services to do translation, or to increase our reach in distance
education or people who have English as a second language. So there's a whole range
of things that improve the business of education. And we're really seeing that
all the way from primary schools all the way through to, you know,
higher degrees in universities.
- Right. And if I'm an IT administrator or someone of an IT
decision-maker background, if I'm having a conversation with, say
a principal or even a vice chancellor, would you suggest then there is a business
justification for starting to adopt some of these technologies or that potential
does exist there?
- Absolutely. And I think, you know, my advice would be to start with something
very simple that people can relate to. So there's nearly always a process or
something in a school or a university or college environment that is causing pain
at the moment. And it might be one of those things I mentioned earlier,
like the fact that somebody somewhere in the school has to spend a lot of time
answering questions about things that people could find on the web,
but they don't. So it might be, perhaps, applying a bot. And the way our bots work,
if you go to the Azure site and you look for Bot Framework, or just search
for "Microsoft Bot Framework," it's very easy to get going. And maybe just start
with a bunch of documents and build a bot that you could use, and demonstrate very
easily to a school principal or an administrator to show how you could
save time and money by using a bot. So it's very approachable.
And starting with practical examples like that before you go to the rocket science
stuff will show people practically how those things can be used.
- And in terms of your suggestion for who would take on that task of building a bot,
trialing a cognitive service in a classroom setting or even a
university setting, do you believe that that would largely depend on the IT
administrator or the IT decision-maker within that institution?
- So not necessarily. In fact, there is certainly a number of really
valuable things that IT can use those technologies for themselves to improve the
IT services in a school or university. But we're actually seeing that teachers
and even people who are not necessarily super technology literate… For example,
the bot example I mentioned earlier, if you can use copy and paste
on a computer, you can create a bot. And we are literally seeing that more and
more teachers and students are starting to discover that not only are these services
really amazing in terms of what they can achieve, but they're super-easy
to get into. They're very approachable. As I said, if you can use copy and paste,
if you can use a mouse, if you can go to a website,
you can start using these services. And that's one of the things we've tried
very hard to do at Microsoft, is to make sure that these services are
not only powerful, but they're also approachable. Because if they're not
usable by everyday people, you're not going to get the uptake.
So, yeah, it can be either/or.
- Right. And say I'm ready to embark on this journey or find out more information.
Is this a free service? How is it licensed? Where can I get my
hands on the trial or a bit of on-demand training? Do you have any recommendations
or suggestions for that?
- Sure. So, when it comes to the Cognitive Services or bots, first,
we recommend that you go and you search for "Azure Cognitive Services," and it'll
bring up the site, the Microsoft site around that. And that has all the
information you need, everything to get started,
the kits and so on. Most of the tools that you need are either online or down...
- So if I'm someone who wants to get started with Cognitive Services,
whereabouts would I go and what is it that I need to know about the pricing model or
what's available for me to trial?
- Sure. So a couple of things to bear in mind. Firstly, the tools to build
cognitive services, to build with Cognitive Services, and to build bots and
to do that sort of work, most of those tools are free.
So if you search up, "Azure Cognitive Services," or, "Azure Bot Framework,"
you'll find the two landing pages for those two services, and you'll be able to
read all about them. So getting into and using the tools is free.
The way that these services
actually work is you only pay for what you use. So for example, if you build a bot,
you're only going to pay for the time that somebody is actually using that bot.
And what you find is, as with most cloud services,
the charges are really pretty small. And unless you're doing something on a
massive scale with 100,000 users or a million users, well, yeah,
then obviously the cost goes up. But for getting started,
you'll find that the cost for those things is very small.
Secondly, if you go and search for, "Microsoft Imagine,"
you'll find that we have a set of tools and resources for educators to use that
will provide you free access to some of these resources in Azure.
So if you want to work with your students on helping to build and learn
about these technologies, that's free. So you can get started completely for free
and start to use some of those services with your students.
- Well, that's great, Mark. And any last words of wisdom
for our viewers? Is there a cognitive service, for example,
that is a favorite of yours that you believe people should have a go on?
- Sure. So certainly, I would start with the Bot Framework
because that's really great. If you've literally got like an FAQ
document or something that you use, really easy place to start. Build a bot,
copy and paste the questions from the document into the Bot Framework,
which you can do online just using a web browser, and you will straight away in a
few minutes be able to see how those work. Once you get excited about that,
there are all sorts of things you can throw at it. The second one that is a
personal favorite of mine is the image recognition service. So what that does,
it'll be able to look at an image and determine using AI what's in the image.
So what is it? Is it a person, a car, an object? I personally find it fun
to play with that one and see what we can do to understand what's inside an image.
So yeah, that would be a couple of places I'd start.
- Right. Well, thank you so much, Mark. Thanks for spending the time with us.
- Thank you. Appreciate it.
Episode 3: IT IQ Series. Episode 1: IT IQ Series. Goodments are using Microsoft Azure to help you invest to make money, and do good Episode 2: IT IQ Series. Episode 2: Using OneNote Class Notebook within Microsoft Teams Episode 3: Embedding External Content and the MEC Episode 1: Introduction to Microsoft Teams Bricks + Agent leverage Microsoft Azure to provide all your property services Mashay harnesses the power of Microsoft Azure to deliver unique experiences Introducing Surface Laptop. Powerfully Beautiful

Episode 4: IT IQ Series.

Latest Posts