Up next in 10
Join this live session with Magnus Mårtensson ft. Foyin Olajide Bello for the next episode of The Cloud Show with Magnus Mårtensson on June 12, at 01:05 PM (EST).
The show is about cloud leadership and all the important questions relating to cloud projects. Certainly, many matters when a company is going to and wants to be successful in the cloud, are about technology. However, there are many additional matters, adjacent to technology, that we also need to tend to regarding business strategy, human resources, organizational change, planning for a technical cloud approach, and many more questions. These conversations are critical for a healthy cloud and for a swift and accurate cloud approach.
GUEST SPEAKER
Foyin has a deep "love" for Microsoft's Power Platform and she has a wide experience structuring and implementing impactful organizational projects with these tools. She is passionate about this because it empowers people to do MORE with less.
She strongly believes in knowledge sharing hence leads an active Power Platform user group in Nigeria. She is also an advocate of women empowerment through Information Technology and has recently founded TechStylers – a community to upskill women. She speaks frequently at technology events and is an active member in similar international groups.
📺 CSharp TV - Dev Streaming Destination http://csharp.tv
🌎 C# Corner - Community of Software and Data Developers https://www.c-sharpcorner.com/
#CSharpTV #csharpcorner #TheCloudShow #CSharp #Interview
Transcript
Show More Show Less View Video Transcript
0:00
[Music]
0:10
hello everyone and welcome back again to
0:12
hello everyone and welcome back again to
0:12
hello everyone and welcome back again to the cloud show we have another exciting
0:15
the cloud show we have another exciting
0:15
the cloud show we have another exciting episode today I'm going to talk to
0:18
episode today I'm going to talk to
0:18
episode today I'm going to talk to foell this is a difficult name I
0:20
foell this is a difficult name I
0:20
foell this is a difficult name I apologize she is a Solutions architect
0:23
apologize she is a Solutions architect
0:24
apologize she is a Solutions architect with aanad Ireland and um we she's a an
0:28
with aanad Ireland and um we she's a an
0:28
with aanad Ireland and um we she's a an expert in Ai and that's what we're going
0:30
expert in Ai and that's what we're going
0:30
expert in Ai and that's what we're going to talk about today which is going to be
0:31
to talk about today which is going to be
0:31
to talk about today which is going to be so fascinating so um as an expert in AI
0:35
so fascinating so um as an expert in AI
0:35
so fascinating so um as an expert in AI you get to explain AI to people and try
0:37
you get to explain AI to people and try
0:37
you get to explain AI to people and try to make people understand what AI is and
0:40
to make people understand what AI is and
0:40
to make people understand what AI is and what it can do so explainable AI has
0:42
what it can do so explainable AI has
0:42
what it can do so explainable AI has some emerging emerging challenges to it
0:46
some emerging emerging challenges to it
0:46
some emerging emerging challenges to it and that is what I will be covering
0:48
and that is what I will be covering
0:48
and that is what I will be covering today with the guest star of the show
0:50
today with the guest star of the show
0:50
today with the guest star of the show fing welcome
1:02
hello
1:02
hello
1:03
hello F hi hi very welcome to the show it's
1:07
F hi hi very welcome to the show it's
1:07
F hi hi very welcome to the show it's nice to have you with
1:09
nice to have you with
1:09
nice to have you with us thank you so much for inviting me
1:13
us thank you so much for inviting me
1:13
us thank you so much for inviting me absolutely my pleasure definitely so AI
1:16
absolutely my pleasure definitely so AI
1:16
absolutely my pleasure definitely so AI is is a wonderful topic and before we
1:18
is is a wonderful topic and before we
1:18
is is a wonderful topic and before we dive into that I want to know a little
1:20
dive into that I want to know a little
1:20
dive into that I want to know a little bit more about you and about you work
1:23
bit more about you and about you work
1:23
bit more about you and about you work you work at aanad yes I do in Ireland
1:27
you work at aanad yes I do in Ireland
1:27
you work at aanad yes I do in Ireland yes in yeah out of Dublin
1:30
yes in yeah out of Dublin
1:30
yes in yeah out of Dublin based out of Dublin I have never been to
1:31
based out of Dublin I have never been to
1:32
based out of Dublin I have never been to Dublin I hope to get to go someday it
1:33
Dublin I hope to get to go someday it
1:33
Dublin I hope to get to go someday it will be fun ah nice nice it's it's
1:36
will be fun ah nice nice it's it's
1:36
will be fun ah nice nice it's it's really lovely yeah lovely hair and the
1:38
really lovely yeah lovely hair and the
1:38
really lovely yeah lovely hair and the lovely weather too right and as as a
1:42
lovely weather too right and as as a
1:42
lovely weather too right and as as a Solutions architect tell us what is it
1:44
Solutions architect tell us what is it
1:44
Solutions architect tell us what is it that you do on a on a daily
1:47
that you do on a on a daily
1:47
that you do on a on a daily basis yes so as a solution architect my
1:50
basis yes so as a solution architect my
1:50
basis yes so as a solution architect my um main goal is to help and support our
1:54
um main goal is to help and support our
1:54
um main goal is to help and support our clients um along their digital
1:56
clients um along their digital
1:56
clients um along their digital transformation Journey so this cuts
1:59
transformation Journey so this cuts
1:59
transformation Journey so this cuts across
2:00
across
2:00
across various tools that we use today from
2:04
various tools that we use today from
2:04
various tools that we use today from Microsoft 365 in its entirety to the
2:07
Microsoft 365 in its entirety to the
2:07
Microsoft 365 in its entirety to the Power Platform to Microsoft 365
2:10
Power Platform to Microsoft 365
2:10
Power Platform to Microsoft 365 co-pilots to Viva just basically
2:13
co-pilots to Viva just basically
2:13
co-pilots to Viva just basically ensuring that whatever like whatever it
2:16
ensuring that whatever like whatever it
2:16
ensuring that whatever like whatever it is that our clients are trying to
2:18
is that our clients are trying to
2:18
is that our clients are trying to achieve um I'm supporting them on that
2:21
achieve um I'm supporting them on that
2:21
achieve um I'm supporting them on that journey and making sure that they achiev
2:23
journey and making sure that they achiev
2:23
journey and making sure that they achiev that and even
2:24
that and even
2:24
that and even more in the best way of course of course
2:28
more in the best way of course of course
2:28
more in the best way of course of course yeah yeah know that I'm sure that that
2:30
yeah yeah know that I'm sure that that
2:30
yeah yeah know that I'm sure that that is a a great big lot and then very
2:33
is a a great big lot and then very
2:33
is a a great big lot and then very recently everything just had made a a
2:36
recently everything just had made a a
2:36
recently everything just had made a a huge pivot uh everything just you know
2:40
huge pivot uh everything just you know
2:40
huge pivot uh everything just you know if you're not talking about AI you're
2:41
if you're not talking about AI you're
2:41
if you're not talking about AI you're not doing it right right I know I know
2:45
not doing it right right I know I know
2:46
not doing it right right I know I know right now everyone has to talk about ai
2:48
right now everyone has to talk about ai
2:48
right now everyone has to talk about ai ai is the thing and so you as an expert
2:52
ai is the thing and so you as an expert
2:52
ai is the thing and so you as an expert uh will as well be in the same situation
2:54
uh will as well be in the same situation
2:54
uh will as well be in the same situation you have to now talk about
2:56
you have to now talk about
2:56
you have to now talk about AI yes 100% And it's interesting I was
3:01
AI yes 100% And it's interesting I was
3:01
AI yes 100% And it's interesting I was in a workshop a while ago and I was
3:03
in a workshop a while ago and I was
3:03
in a workshop a while ago and I was saying to I was saying to my audience AI
3:06
saying to I was saying to my audience AI
3:06
saying to I was saying to my audience AI didn't knew you know I think the first
3:08
didn't knew you know I think the first
3:08
didn't knew you know I think the first time I wor with AI was almost when
3:11
time I wor with AI was almost when
3:11
time I wor with AI was almost when 2012 when there was the um we had a
3:15
2012 when there was the um we had a
3:15
2012 when there was the um we had a model for optical character recognition
3:18
model for optical character recognition
3:18
model for optical character recognition and we wanted to like scan the forms and
3:20
and we wanted to like scan the forms and
3:20
and we wanted to like scan the forms and pick out certain values from certain
3:22
pick out certain values from certain
3:22
pick out certain values from certain places and throw it into a DB um and you
3:25
places and throw it into a DB um and you
3:25
places and throw it into a DB um and you know I think it's it's the Gen that's
3:28
know I think it's it's the Gen that's
3:28
know I think it's it's the Gen that's really really caught the the big
3:30
really really caught the the big
3:30
really really caught the the big attention now of everyone certainly
3:33
attention now of everyone certainly
3:33
attention now of everyone certainly something that we need to talk about and
3:34
something that we need to talk about and
3:34
something that we need to talk about and keep talking about yeah AI is hardly new
3:37
keep talking about yeah AI is hardly new
3:37
keep talking about yeah AI is hardly new we were talking about AI or there were
3:39
we were talking about AI or there were
3:39
we were talking about AI or there were movies about AI taking over the world in
3:41
movies about AI taking over the world in
3:41
movies about AI taking over the world in the 80s or something like that right
3:43
the 80s or something like that right
3:43
the 80s or something like that right this is I mean this is not at all
3:45
this is I mean this is not at all
3:45
this is I mean this is not at all anything new but what is it's
3:47
anything new but what is it's
3:47
anything new but what is it's fascinating that it's caught on in such
3:49
fascinating that it's caught on in such
3:49
fascinating that it's caught on in such a huge um you know popular wave right
3:54
a huge um you know popular wave right
3:54
a huge um you know popular wave right now do you have any idea why this
3:56
now do you have any idea why this
3:56
now do you have any idea why this is I think it's I think the the reason
3:59
is I think it's I think the the reason
3:59
is I think it's I think the the reason why it's cutting such it's getting such
4:01
why it's cutting such it's getting such
4:01
why it's cutting such it's getting such a great wave is the whole gen AI the
4:05
a great wave is the whole gen AI the
4:05
a great wave is the whole gen AI the fact that it's it retains context I
4:08
fact that it's it retains context I
4:08
fact that it's it retains context I think that's the big thing as the user
4:11
think that's the big thing as the user
4:11
think that's the big thing as the user started to get use um tools like chart
4:14
started to get use um tools like chart
4:14
started to get use um tools like chart GTP and the fact that it could kind of
4:16
GTP and the fact that it could kind of
4:17
GTP and the fact that it could kind of answer your questions in a very
4:20
answer your questions in a very
4:20
answer your questions in a very interesting and different way and keep
4:22
interesting and different way and keep
4:22
interesting and different way and keep the context of that
4:24
the context of that
4:24
the context of that conversation um as you go along so I
4:27
conversation um as you go along so I
4:27
conversation um as you go along so I would say that that really created that
4:30
would say that that really created that
4:30
would say that that really created that differential for people with it yeah
4:34
differential for people with it yeah
4:34
differential for people with it yeah that makes a lot of sense to me as well
4:37
that makes a lot of sense to me as well
4:37
that makes a lot of sense to me as well I I definitely agree so the fact that um
4:40
I I definitely agree so the fact that um
4:40
I I definitely agree so the fact that um you can appear to have a conversation
4:43
you can appear to have a conversation
4:43
you can appear to have a conversation with it and it it it remembers what you
4:45
with it and it it it remembers what you
4:45
with it and it it it remembers what you just said and responds accordingly it go
4:48
just said and responds accordingly it go
4:48
just said and responds accordingly it go it's it's a back and forth thing as
4:50
it's it's a back and forth thing as
4:50
it's it's a back and forth thing as compared to just writing a question into
4:52
compared to just writing a question into
4:52
compared to just writing a question into get some good results but you know it's
4:55
get some good results but you know it's
4:55
get some good results but you know it's you ask the next question it doesn't
4:56
you ask the next question it doesn't
4:56
you ask the next question it doesn't remember what you just said exactly so
4:59
remember what you just said exactly so
4:59
remember what you just said exactly so is that conversational AI that context
5:02
is that conversational AI that context
5:02
is that conversational AI that context that was the big um differential in this
5:06
that was the big um differential in this
5:06
that was the big um differential in this one and everyone just kind of jumped on
5:08
one and everyone just kind of jumped on
5:08
one and everyone just kind of jumped on it because was really cool that's
5:11
it because was really cool that's
5:12
it because was really cool that's absolutely interesting I I'm trying to
5:15
absolutely interesting I I'm trying to
5:15
absolutely interesting I I'm trying to figure this out in my brain as you say
5:18
figure this out in my brain as you say
5:18
figure this out in my brain as you say it it sounds like you're incredibly spot
5:21
it it sounds like you're incredibly spot
5:21
it it sounds like you're incredibly spot on now so given given that we are now
5:24
on now so given given that we are now
5:24
on now so given given that we are now having something like a conversation
5:26
having something like a conversation
5:26
having something like a conversation with an AI and and we are using it in in
5:29
with an AI and and we are using it in in
5:29
with an AI and and we are using it in in such a massive more um you know all
5:33
such a massive more um you know all
5:33
such a massive more um you know all businesses want to have something AI
5:35
businesses want to have something AI
5:35
businesses want to have something AI it's like it's essentially whatever
5:37
it's like it's essentially whatever
5:37
it's like it's essentially whatever product we had before now with AI right
5:41
product we had before now with AI right
5:41
product we had before now with AI right AI is a solution and companies are going
5:44
AI is a solution and companies are going
5:44
AI is a solution and companies are going out to try to find the problem um so so
5:48
out to try to find the problem um so so
5:48
out to try to find the problem um so so what do you do to advise your clients
5:52
what do you do to advise your clients
5:52
what do you do to advise your clients about AI
5:54
about AI
5:54
about AI now yeah I think so our our Approach at
5:58
now yeah I think so our our Approach at
5:58
now yeah I think so our our Approach at avanar is first of all AI first means
6:00
avanar is first of all AI first means
6:00
avanar is first of all AI first means people first that's one of the important
6:02
people first that's one of the important
6:02
people first that's one of the important principles that we have that is
6:04
principles that we have that is
6:04
principles that we have that is governing us um as an organization and
6:07
governing us um as an organization and
6:07
governing us um as an organization and one of the things that we we help our
6:09
one of the things that we we help our
6:09
one of the things that we we help our clients to do is ensure that the
6:12
clients to do is ensure that the
6:12
clients to do is ensure that the foundations are are set right right for
6:15
foundations are are set right right for
6:15
foundations are are set right right for example if you want to get started with
6:17
example if you want to get started with
6:17
example if you want to get started with M 365 co-pilot you need to ensure that
6:20
M 365 co-pilot you need to ensure that
6:20
M 365 co-pilot you need to ensure that some housekeeping is in place right like
6:23
some housekeeping is in place right like
6:23
some housekeeping is in place right like your SharePoint sharing sharing of
6:26
your SharePoint sharing sharing of
6:26
your SharePoint sharing sharing of documents on your SharePoint one drive
6:28
documents on your SharePoint one drive
6:29
documents on your SharePoint one drive like over sharing needs to be taken care
6:30
like over sharing needs to be taken care
6:30
like over sharing needs to be taken care of so that ensures that once Microsoft
6:33
of so that ensures that once Microsoft
6:33
of so that ensures that once Microsoft 365 is in use and people start searching
6:36
365 is in use and people start searching
6:36
365 is in use and people start searching they don't start seeing what they
6:37
they don't start seeing what they
6:37
they don't start seeing what they shouldn't be seeing or accessing the
6:39
shouldn't be seeing or accessing the
6:39
shouldn't be seeing or accessing the documents that it shouldn't access so
6:41
documents that it shouldn't access so
6:41
documents that it shouldn't access so Microsoft doesn't um Microsoft doesn't
6:45
Microsoft doesn't um Microsoft doesn't
6:45
Microsoft doesn't um Microsoft doesn't take our data but all of the security
6:48
take our data but all of the security
6:48
take our data but all of the security around a tool like myself 365 is relying
6:52
around a tool like myself 365 is relying
6:52
around a tool like myself 365 is relying heavily on the existing you know
6:54
heavily on the existing you know
6:54
heavily on the existing you know security and data structure in place in
6:57
security and data structure in place in
6:57
security and data structure in place in the organization so a lot of
6:58
the organization so a lot of
6:58
the organization so a lot of organizations
7:00
organizations
7:00
organizations haven't gotten that right yet but they
7:03
haven't gotten that right yet but they
7:03
haven't gotten that right yet but they want to start using a so it's about
7:05
want to start using a so it's about
7:05
want to start using a so it's about ensuring that all of those things are in
7:07
ensuring that all of those things are in
7:07
ensuring that all of those things are in place and they putting AI on top of that
7:11
place and they putting AI on top of that
7:11
place and they putting AI on top of that ahuh that that's that's an important
7:14
ahuh that that's that's an important
7:14
ahuh that that's that's an important area because so technically when a
7:17
area because so technically when a
7:17
area because so technically when a company wants to use AI they're mostly
7:20
company wants to use AI they're mostly
7:20
company wants to use AI they're mostly interested in using it together with
7:22
interested in using it together with
7:22
interested in using it together with their own data they have a lot of data
7:25
their own data they have a lot of data
7:25
their own data they have a lot of data and if only we could have a conversation
7:27
and if only we could have a conversation
7:27
and if only we could have a conversation with that data if only we can do
7:28
with that data if only we can do
7:28
with that data if only we can do something really smart with AI with all
7:31
something really smart with AI with all
7:31
something really smart with AI with all the massive amounts of data that we have
7:33
the massive amounts of data that we have
7:33
the massive amounts of data that we have um now you have to take care of that
7:35
um now you have to take care of that
7:35
um now you have to take care of that data in such a way that you you take it
7:37
data in such a way that you you take it
7:37
data in such a way that you you take it and make it available for the AI to
7:40
and make it available for the AI to
7:40
and make it available for the AI to train the AI with your data is that is
7:42
train the AI with your data is that is
7:42
train the AI with your data is that is that so far appropriate yes yes it's two
7:46
that so far appropriate yes yes it's two
7:46
that so far appropriate yes yes it's two it's two things it's either you buy or
7:48
it's two things it's either you buy or
7:48
it's two things it's either you buy or you build so you can buy tools like my
7:51
you build so you can buy tools like my
7:51
you build so you can buy tools like my 365 co-pilot or any of the co-pilots
7:54
365 co-pilot or any of the co-pilots
7:54
365 co-pilot or any of the co-pilots co-pilot for security co-pilot for sales
7:57
co-pilot for security co-pilot for sales
7:57
co-pilot for security co-pilot for sales in Dynamics in Power Platform Copilot is
7:59
in Dynamics in Power Platform Copilot is
7:59
in Dynamics in Power Platform Copilot is present everywhere um in all of
8:01
present everywhere um in all of
8:01
present everywhere um in all of Microsoft PS5 apps in all of Microsoft's
8:05
Microsoft PS5 apps in all of Microsoft's
8:05
Microsoft PS5 apps in all of Microsoft's app applications or you say you know
8:08
app applications or you say you know
8:08
app applications or you say you know what I want something different
8:10
what I want something different
8:10
what I want something different something unique and then you can build
8:12
something unique and then you can build
8:12
something unique and then you can build your own leveraging some of the models
8:15
your own leveraging some of the models
8:15
your own leveraging some of the models available in Azure yeah right so you can
8:18
available in Azure yeah right so you can
8:18
available in Azure yeah right so you can build your own AI you can build your own
8:20
build your own AI you can build your own
8:20
build your own AI you can build your own co-pilots but you can also use the ones
8:22
co-pilots but you can also use the ones
8:22
co-pilots but you can also use the ones that exist there are so many that nobody
8:25
that exist there are so many that nobody
8:25
that exist there are so many that nobody can count them uh but but given that you
8:28
can count them uh but but given that you
8:28
can count them uh but but given that you want to take them and point them to your
8:30
want to take them and point them to your
8:30
want to take them and point them to your company's data and and there there where
8:32
company's data and and there there where
8:32
company's data and and there there where I understand that you have to first go
8:34
I understand that you have to first go
8:34
I understand that you have to first go in and ensure that your that the company
8:37
in and ensure that your that the company
8:37
in and ensure that your that the company the client is not
8:39
the client is not
8:39
the client is not oversharing uh data or they're sharing
8:42
oversharing uh data or they're sharing
8:42
oversharing uh data or they're sharing the appropriate
8:44
the appropriate
8:44
the appropriate data making it available to the AI but
8:46
data making it available to the AI but
8:46
data making it available to the AI but not data that they don't want the AI to
8:49
not data that they don't want the AI to
8:49
not data that they don't want the AI to know about right yes yes so it's just
8:52
know about right yes yes so it's just
8:52
know about right yes yes so it's just important yeah those are I mean that's
8:54
important yeah those are I mean that's
8:54
important yeah those are I mean that's one of the many things that needs to be
8:56
one of the many things that needs to be
8:56
one of the many things that needs to be done but that's that's what we've seen
8:57
done but that's that's what we've seen
8:57
done but that's that's what we've seen as the most well one of the the most
8:59
as the most well one of the the most
8:59
as the most well one of the the most important things you need to ensure
9:01
important things you need to ensure
9:01
important things you need to ensure before you start to enable tools like
9:03
before you start to enable tools like
9:03
before you start to enable tools like Microsoft 365
9:04
Microsoft 365
9:04
Microsoft 365 co-pilots yeah yeah yeah because you
9:07
co-pilots yeah yeah yeah because you
9:07
co-pilots yeah yeah yeah because you don't want maybe anyone to start
9:10
don't want maybe anyone to start
9:10
don't want maybe anyone to start suddenly reading the Personnel files and
9:13
suddenly reading the Personnel files and
9:13
suddenly reading the Personnel files and you know you want to make sure that
9:15
you know you want to make sure that
9:15
you know you want to make sure that those things are you know
9:17
those things are you know
9:17
those things are you know appropriately available to only the HR
9:19
appropriately available to only the HR
9:19
appropriately available to only the HR department Pro probably right because
9:22
department Pro probably right because
9:22
department Pro probably right because otherwise things can go really really
9:23
otherwise things can go really really
9:23
otherwise things can go really really bad yeah another important thing is
9:26
bad yeah another important thing is
9:26
bad yeah another important thing is responsible AI um and like for us at
9:30
responsible AI um and like for us at
9:30
responsible AI um and like for us at aard we ensure that everyone was trained
9:33
aard we ensure that everyone was trained
9:33
aard we ensure that everyone was trained um in responsible a and what that means
9:35
um in responsible a and what that means
9:35
um in responsible a and what that means for us as an organization so there's
9:37
for us as an organization so there's
9:37
for us as an organization so there's education element as well because like
9:40
education element as well because like
9:40
education element as well because like organizations can say oh no we're not
9:43
organizations can say oh no we're not
9:43
organizations can say oh no we're not we're not going to use AI in
9:44
we're not going to use AI in
9:45
we're not going to use AI in organization that's fine but the truth
9:47
organization that's fine but the truth
9:47
organization that's fine but the truth is you never can tell if your employees
9:49
is you never can tell if your employees
9:49
is you never can tell if your employees are going to charge GTP right to find
9:51
are going to charge GTP right to find
9:51
are going to charge GTP right to find information on their own after work
9:53
information on their own after work
9:53
information on their own after work hours with their on their own PCS so
9:55
hours with their on their own PCS so
9:55
hours with their on their own PCS so it's important to get everyone trained
9:58
it's important to get everyone trained
9:58
it's important to get everyone trained and help them understand what
10:00
and help them understand what
10:00
and help them understand what responsible AI means for your
10:02
responsible AI means for your
10:02
responsible AI means for your organization what ethical AI means for
10:04
organization what ethical AI means for
10:04
organization what ethical AI means for your organization what is appropriate
10:06
your organization what is appropriate
10:06
your organization what is appropriate and what is not um so that helps them to
10:11
and what is not um so that helps them to
10:11
and what is not um so that helps them to understand okay this is what this is how
10:13
understand okay this is what this is how
10:13
understand okay this is what this is how our organization is approaching the use
10:15
our organization is approaching the use
10:15
our organization is approaching the use of AI in our business and then they can
10:19
of AI in our business and then they can
10:19
of AI in our business and then they can you know use it appropriately so
10:21
you know use it appropriately so
10:21
you know use it appropriately so creating that guidance is also important
10:23
creating that guidance is also important
10:23
creating that guidance is also important otherwise we're leaving the employees in
10:25
otherwise we're leaving the employees in
10:25
otherwise we're leaving the employees in the workplace to kind of figure it out
10:26
the workplace to kind of figure it out
10:26
the workplace to kind of figure it out themselves and as you know ethic
10:30
themselves and as you know ethic
10:30
themselves and as you know ethic people can Define ethics in very
10:32
people can Define ethics in very
10:32
people can Define ethics in very different ways yes yes they can indeed
10:34
different ways yes yes they can indeed
10:35
different ways yes yes they can indeed and and so I'm thinking about all those
10:36
and and so I'm thinking about all those
10:36
and and so I'm thinking about all those those you know this is a huge topic area
10:39
those you know this is a huge topic area
10:39
those you know this is a huge topic area I'm sure you can talk about this as an
10:41
I'm sure you can talk about this as an
10:41
I'm sure you can talk about this as an expert I'm very sure you can talk about
10:43
expert I'm very sure you can talk about
10:43
expert I'm very sure you can talk about this for days um given just a few
10:46
this for days um given just a few
10:46
this for days um given just a few moments or or the the the length of this
10:48
moments or or the the the length of this
10:48
moments or or the the the length of this conversation ethical and and responsible
10:52
conversation ethical and and responsible
10:52
conversation ethical and and responsible AI for the learner what what is this
10:56
AI for the learner what what is this
10:56
AI for the learner what what is this like more what what's it it about
11:00
like more what what's it it about
11:00
like more what what's it it about really so if you're being unethical or
11:03
really so if you're being unethical or
11:03
really so if you're being unethical or if you're not being responsible what
11:05
if you're not being responsible what
11:05
if you're not being responsible what happens
11:06
happens
11:06
happens then well that then depends on the
11:09
then well that then depends on the
11:09
then well that then depends on the policy of your organization so it's
11:11
policy of your organization so it's
11:12
policy of your organization so it's mainly around three three things I would
11:15
mainly around three three things I would
11:15
mainly around three three things I would say the first thing is the principles
11:17
say the first thing is the principles
11:17
say the first thing is the principles there are principles around respon the
11:19
there are principles around respon the
11:19
there are principles around respon the responsible use of AI and ethical use of
11:21
responsible use of AI and ethical use of
11:21
responsible use of AI and ethical use of AI so example making sure that it's bias
11:24
AI so example making sure that it's bias
11:24
AI so example making sure that it's bias free ensuring that it's trustworthy it's
11:27
free ensuring that it's trustworthy it's
11:27
free ensuring that it's trustworthy it's reliable there's accountability
11:29
reliable there's accountability
11:29
reliable there's accountability place all of that and then there's the
11:33
place all of that and then there's the
11:33
place all of that and then there's the practices so you have the principles but
11:35
practices so you have the principles but
11:35
practices so you have the principles but how do we um actualize this how do we
11:38
how do we um actualize this how do we
11:38
how do we um actualize this how do we bring this principles to life so you
11:40
bring this principles to life so you
11:40
bring this principles to life so you start to talk about the organization
11:42
start to talk about the organization
11:42
start to talk about the organization having an AI ethics board um the
11:46
having an AI ethics board um the
11:46
having an AI ethics board um the organization having um the developers
11:48
organization having um the developers
11:48
organization having um the developers having the ethics by Design principle
11:51
having the ethics by Design principle
11:52
having the ethics by Design principle where ethics doesn't come as an
11:54
where ethics doesn't come as an
11:54
where ethics doesn't come as an aftermath it's kind of right there
11:55
aftermath it's kind of right there
11:55
aftermath it's kind of right there embedded in their thought pattern as
11:58
embedded in their thought pattern as
11:58
embedded in their thought pattern as they start to build and develop velop um
12:00
they start to build and develop velop um
12:00
they start to build and develop velop um the AI models um what else you you talk
12:04
the AI models um what else you you talk
12:04
the AI models um what else you you talk about monitoring so how do we monitor
12:06
about monitoring so how do we monitor
12:06
about monitoring so how do we monitor how do we check that whatever um ai2
12:11
how do we check that whatever um ai2
12:11
how do we check that whatever um ai2 were developing is actually meeting this
12:13
were developing is actually meeting this
12:13
were developing is actually meeting this standard so those are some of the
12:14
standard so those are some of the
12:14
standard so those are some of the practices that you can bring and then we
12:16
practices that you can bring and then we
12:16
practices that you can bring and then we have the standards so the standards are
12:18
have the standards so the standards are
12:18
have the standards so the standards are all of the policies that we have in
12:21
all of the policies that we have in
12:21
all of the policies that we have in place um around like from ISO from I um
12:26
place um around like from ISO from I um
12:26
place um around like from ISO from I um that starts to say okay this is how what
12:29
that starts to say okay this is how what
12:29
that starts to say okay this is how what this is what privacy means this is what
12:31
this is what privacy means this is what
12:31
this is what privacy means this is what um this the um gen use of AI policy
12:35
um this the um gen use of AI policy
12:35
um this the um gen use of AI policy means we also have the EU act around AI
12:38
means we also have the EU act around AI
12:38
means we also have the EU act around AI so those are the standards that then
12:40
so those are the standards that then
12:40
so those are the standards that then help organizations to shape their
12:42
help organizations to shape their
12:42
help organizations to shape their internal policies so as as an
12:45
internal policies so as as an
12:45
internal policies so as as an organization you want to make sure that
12:47
organization you want to make sure that
12:47
organization you want to make sure that you uh abide by the standards and and
12:50
you uh abide by the standards and and
12:50
you uh abide by the standards and and use them in the appropriate way so that
12:52
use them in the appropriate way so that
12:52
use them in the appropriate way so that you can you can be ethical and and
12:55
you can you can be ethical and and
12:55
you can you can be ethical and and responsible about things so for example
12:57
responsible about things so for example
12:57
responsible about things so for example I'm I'm from Sweden and we always like
12:58
I'm I'm from Sweden and we always like
12:58
I'm I'm from Sweden and we always like to make jokes about our neighboring
13:00
to make jokes about our neighboring
13:00
to make jokes about our neighboring country Norway and they they make jokes
13:02
country Norway and they they make jokes
13:02
country Norway and they they make jokes about us but if I have a if I have a
13:05
about us but if I have a if I have a
13:05
about us but if I have a if I have a data source which is heavily biased
13:07
data source which is heavily biased
13:07
data source which is heavily biased against Norwegian people and I I put
13:10
against Norwegian people and I I put
13:10
against Norwegian people and I I put that AI available to to the world um is
13:13
that AI available to to the world um is
13:13
that AI available to to the world um is going to say bad things about Norway uh
13:16
going to say bad things about Norway uh
13:16
going to say bad things about Norway uh which is you know that's not friendly uh
13:18
which is you know that's not friendly uh
13:18
which is you know that's not friendly uh it's maybe not e ethical or or or good
13:20
it's maybe not e ethical or or or good
13:21
it's maybe not e ethical or or or good or maybe not even the intended uh way
13:23
or maybe not even the intended uh way
13:23
or maybe not even the intended uh way but you have to understand that if you
13:25
but you have to understand that if you
13:25
but you have to understand that if you train your data with bias or your AI
13:29
train your data with bias or your AI
13:29
train your data with bias or your AI with biased data your AI will be biased
13:36
100% so there's so many things here to
13:39
100% so there's so many things here to
13:39
100% so there's so many things here to to talk about uh about AI I'm I'm I'm
13:42
to talk about uh about AI I'm I'm I'm
13:42
to talk about uh about AI I'm I'm I'm very glad that we we get to do this but
13:44
very glad that we we get to do this but
13:44
very glad that we we get to do this but so um fixing the data fixing the data
13:48
so um fixing the data fixing the data
13:48
so um fixing the data fixing the data sounds like an easy thing right whatever
13:50
sounds like an easy thing right whatever
13:50
sounds like an easy thing right whatever fixing the data and making sure we're
13:52
fixing the data and making sure we're
13:52
fixing the data and making sure we're behaving uh appropriately um and and
13:56
behaving uh appropriately um and and
13:56
behaving uh appropriately um and and without bias and and and ethically
13:59
without bias and and and ethically
13:59
without bias and and and ethically very important areas and then then what
14:01
very important areas and then then what
14:01
very important areas and then then what we do so you you want to send a you want
14:04
we do so you you want to send a you want
14:04
we do so you you want to send a you want to get a a client started using AI so
14:07
to get a a client started using AI so
14:07
to get a a client started using AI so what what
14:09
what what
14:09
what what happens oh well then it kind of depends
14:11
happens oh well then it kind of depends
14:11
happens oh well then it kind of depends on where where they want to go and what
14:12
on where where they want to go and what
14:12
on where where they want to go and what their needs are from there and then we
14:15
their needs are from there and then we
14:15
their needs are from there and then we start to define the um engagement
14:18
start to define the um engagement
14:18
start to define the um engagement afterwards yeah so you have to like sit
14:20
afterwards yeah so you have to like sit
14:20
afterwards yeah so you have to like sit down and figure out what is it you are
14:22
down and figure out what is it you are
14:22
down and figure out what is it you are actually going to to use AI for yeah
14:25
actually going to to use AI for yeah
14:25
actually going to to use AI for yeah what problems do you want to solve yeah
14:27
what problems do you want to solve yeah
14:27
what problems do you want to solve yeah and sometimes sometimes may not be the
14:29
and sometimes sometimes may not be the
14:29
and sometimes sometimes may not be the answer it may be automating the process
14:31
answer it may be automating the process
14:31
answer it may be automating the process you know and sometimes um they may need
14:35
you know and sometimes um they may need
14:35
you know and sometimes um they may need to build what than buy because their
14:38
to build what than buy because their
14:38
to build what than buy because their need their needs are very Niche so it
14:41
need their needs are very Niche so it
14:41
need their needs are very Niche so it really depends on what the objective is
14:44
really depends on what the objective is
14:44
really depends on what the objective is yeah are you are you often um at this
14:47
yeah are you are you often um at this
14:47
yeah are you are you often um at this point in in history are you often
14:50
point in in history are you often
14:50
point in in history are you often feeling that you are getting sort of
14:52
feeling that you are getting sort of
14:52
feeling that you are getting sort of vague requests from the customers we
14:55
vague requests from the customers we
14:55
vague requests from the customers we need some AI because we have because we
14:58
need some AI because we have because we
14:58
need some AI because we have because we have to have ai and and they don't
14:59
have to have ai and and they don't
15:00
have to have ai and and they don't really know why right are you no not at
15:04
really know why right are you no not at
15:04
really know why right are you no not at all they are very clear on what they
15:05
all they are very clear on what they
15:05
all they are very clear on what they want and they know they are yes yes so
15:08
want and they know they are yes yes so
15:08
want and they know they are yes yes so far they are and and they know exactly
15:11
far they are and and they know exactly
15:11
far they are and and they know exactly um what it is that they want and where
15:14
um what it is that they want and where
15:14
um what it is that they want and where they want to get to but it's just about
15:16
they want to get to but it's just about
15:16
they want to get to but it's just about shaping it to decide how best can we
15:18
shaping it to decide how best can we
15:18
shaping it to decide how best can we meet the requirements so it sounds
15:20
meet the requirements so it sounds
15:20
meet the requirements so it sounds promising actually I I'm I'm a little
15:22
promising actually I I'm I'm a little
15:22
promising actually I I'm I'm a little bit surprised sir thought you were going
15:24
bit surprised sir thought you were going
15:24
bit surprised sir thought you were going to say that oh my God yes some people
15:26
to say that oh my God yes some people
15:26
to say that oh my God yes some people have no idea
15:29
have no idea
15:29
have no idea well maybe we we have that class those
15:31
well maybe we we have that class those
15:31
well maybe we we have that class those class of people as well but not too many
15:33
class of people as well but not too many
15:33
class of people as well but not too many a lot of them are kind of have a have a
15:35
a lot of them are kind of have a have a
15:35
a lot of them are kind of have a have a relatively solid idea of what they want
15:37
relatively solid idea of what they want
15:37
relatively solid idea of what they want to get through with this so it's all
15:38
to get through with this so it's all
15:39
to get through with this so it's all about being um if I if I can can sort of
15:41
about being um if I if I can can sort of
15:41
about being um if I if I can can sort of conclude what what you have been saying
15:43
conclude what what you have been saying
15:43
conclude what what you have been saying here if I if I can summarize being
15:46
here if I if I can summarize being
15:47
here if I if I can summarize being figuring out the the data access do we
15:49
figuring out the the data access do we
15:49
figuring out the the data access do we share the right data and how do we give
15:51
share the right data and how do we give
15:51
share the right data and how do we give access to the data and do we have the
15:54
access to the data and do we have the
15:54
access to the data and do we have the appropriate data available is it
15:56
appropriate data available is it
15:56
appropriate data available is it unbiased or can we can REM remove bias
16:00
unbiased or can we can REM remove bias
16:00
unbiased or can we can REM remove bias can we make make sure that we are are
16:02
can we make make sure that we are are
16:02
can we make make sure that we are are sharing and using this data ethically
16:04
sharing and using this data ethically
16:04
sharing and using this data ethically and then moving forward with actual the
16:07
and then moving forward with actual the
16:07
and then moving forward with actual the creating or either using a co-pilot or
16:10
creating or either using a co-pilot or
16:10
creating or either using a co-pilot or using an existing AI or creating
16:11
using an existing AI or creating
16:12
using an existing AI or creating something of our own which which then
16:14
something of our own which which then
16:14
something of our own which which then gives us the power of AI in our business
16:17
gives us the power of AI in our business
16:17
gives us the power of AI in our business all right so there is one more thing we
16:18
all right so there is one more thing we
16:19
all right so there is one more thing we wanted to talk about or the thing we
16:20
wanted to talk about or the thing we
16:20
wanted to talk about or the thing we really needed to talk about today
16:22
really needed to talk about today
16:22
really needed to talk about today explainable AI we've talking about
16:24
explainable AI we've talking about
16:24
explainable AI we've talking about ethics and data and many things now you
16:27
ethics and data and many things now you
16:27
ethics and data and many things now you say that there is something called
16:28
say that there is something called
16:28
say that there is something called explainable AI can you explain to us
16:31
explainable AI can you explain to us
16:31
explainable AI can you explain to us what that
16:32
what that
16:32
what that is thank you so um explainable AI is the
16:37
is thank you so um explainable AI is the
16:37
is thank you so um explainable AI is the ability for AI to tell us why it comes
16:40
ability for AI to tell us why it comes
16:40
ability for AI to tell us why it comes up with his conclusions and the answers
16:43
up with his conclusions and the answers
16:43
up with his conclusions and the answers we get so we've seen that that's really
16:45
we get so we've seen that that's really
16:45
we get so we've seen that that's really where people are interested in the
16:47
where people are interested in the
16:47
where people are interested in the software Engineers the developers they
16:49
software Engineers the developers they
16:49
software Engineers the developers they want to know why did you come up with
16:50
want to know why did you come up with
16:50
want to know why did you come up with this answer what's the reasoning behind
16:52
this answer what's the reasoning behind
16:52
this answer what's the reasoning behind it why have you told me this should be
16:54
it why have you told me this should be
16:54
it why have you told me this should be white not black um so that's the real
16:58
white not black um so that's the real
16:58
white not black um so that's the real really where it's going to and it's it's
17:01
really where it's going to and it's it's
17:01
really where it's going to and it's it's easy enough with the decision tree type
17:03
easy enough with the decision tree type
17:03
easy enough with the decision tree type of very simple and straightforward um AI
17:07
of very simple and straightforward um AI
17:07
of very simple and straightforward um AI models but when you start to get to deep
17:08
models but when you start to get to deep
17:08
models but when you start to get to deep learning then it gets complex because
17:11
learning then it gets complex because
17:11
learning then it gets complex because the technology the thinking and behind
17:14
the technology the thinking and behind
17:14
the technology the thinking and behind it is very Advanced and then it starts
17:17
it is very Advanced and then it starts
17:17
it is very Advanced and then it starts to say okay the developers and the
17:19
to say okay the developers and the
17:19
to say okay the developers and the engineers ml Engineers start to think
17:21
engineers ml Engineers start to think
17:21
engineers ml Engineers start to think about how do we explain this in an easy
17:23
about how do we explain this in an easy
17:23
about how do we explain this in an easy enough way for people to understand so
17:26
enough way for people to understand so
17:26
enough way for people to understand so the emerging um RS and Trend here is be
17:32
the emerging um RS and Trend here is be
17:32
the emerging um RS and Trend here is be picking
17:33
picking
17:33
picking between do I make this AI model
17:36
between do I make this AI model
17:36
between do I make this AI model extremely complex um thereby um giving
17:40
extremely complex um thereby um giving
17:40
extremely complex um thereby um giving make making it very efficient for the
17:43
make making it very efficient for the
17:43
make making it very efficient for the users and the B however it's creating
17:45
users and the B however it's creating
17:45
users and the B however it's creating that black box
17:46
that black box
17:46
that black box where it's becoming very difficult to
17:49
where it's becoming very difficult to
17:50
where it's becoming very difficult to explain in simple terms the reasoning
17:52
explain in simple terms the reasoning
17:52
explain in simple terms the reasoning behind the answer it gives or do I make
17:54
behind the answer it gives or do I make
17:54
behind the answer it gives or do I make this model simpler where it can explain
17:57
this model simpler where it can explain
17:57
this model simpler where it can explain but may not give the quality of
17:59
but may not give the quality of
17:59
but may not give the quality of responses that you will get with a
18:01
responses that you will get with a
18:01
responses that you will get with a extremely complex model so there's a
18:03
extremely complex model so there's a
18:03
extremely complex model so there's a potential trade-off in the model
18:05
potential trade-off in the model
18:05
potential trade-off in the model complexity um complexity and
18:08
complexity um complexity and
18:08
complexity um complexity and explainability right right so and so
18:12
explainability right right so and so
18:12
explainability right right so and so explainable
18:14
explainable
18:14
explainable AI what you need to like you need to
18:17
AI what you need to like you need to
18:17
AI what you need to like you need to help the user of the AI understand where
18:21
help the user of the AI understand where
18:21
help the user of the AI understand where the information came from yes it's it's
18:26
the information came from yes it's it's
18:26
the information came from yes it's it's is meeting the challenge of transparency
18:28
is meeting the challenge of transparency
18:28
is meeting the challenge of transparency and expl ability in AI models helping
18:32
and expl ability in AI models helping
18:32
and expl ability in AI models helping the users of the models to be able to
18:34
the users of the models to be able to
18:34
the users of the models to be able to ask why are you thinking this way where
18:36
ask why are you thinking this way where
18:36
ask why are you thinking this way where is this answer coming from what is the
18:37
is this answer coming from what is the
18:37
is this answer coming from what is the reasoning behind this response you have
18:40
reasoning behind this response you have
18:40
reasoning behind this response you have given me so it's it's to eradicate the
18:43
given me so it's it's to eradicate the
18:43
given me so it's it's to eradicate the black box and create that transparency
18:46
black box and create that transparency
18:46
black box and create that transparency where people can understand the
18:48
where people can understand the
18:48
where people can understand the reasoning behind the model um so that's
18:50
reasoning behind the model um so that's
18:50
reasoning behind the model um so that's what explainable AI is but then you have
18:53
what explainable AI is but then you have
18:53
what explainable AI is but then you have the um emerging RS where the more
18:57
the um emerging RS where the more
18:57
the um emerging RS where the more complex the model is especially deep
18:59
complex the model is especially deep
18:59
complex the model is especially deep deep learning models the more difficult
19:01
deep learning models the more difficult
19:01
deep learning models the more difficult it is right to explain in simple terms
19:06
it is right to explain in simple terms
19:06
it is right to explain in simple terms so it's kind of create creating that
19:08
so it's kind of create creating that
19:08
so it's kind of create creating that potential trade of where they can
19:11
potential trade of where they can
19:12
potential trade of where they can developers are trying to decide okay do
19:13
developers are trying to decide okay do
19:13
developers are trying to decide okay do we make it simpler and then it's easy to
19:16
we make it simpler and then it's easy to
19:16
we make it simpler and then it's easy to explain and the transparency is there
19:18
explain and the transparency is there
19:18
explain and the transparency is there 100% or do we go all the way where the
19:20
100% or do we go all the way where the
19:20
100% or do we go all the way where the model is very complex gives us very high
19:22
model is very complex gives us very high
19:22
model is very complex gives us very high quality answers but then it's not really
19:25
quality answers but then it's not really
19:25
quality answers but then it's not really able to explain in simple terms why and
19:27
able to explain in simple terms why and
19:27
able to explain in simple terms why and the reasoning behind it oh wow that's
19:30
the reasoning behind it oh wow that's
19:30
the reasoning behind it oh wow that's such a fascinating topic wow I I think
19:32
such a fascinating topic wow I I think
19:32
such a fascinating topic wow I I think we could talk all evening uh about this
19:35
we could talk all evening uh about this
19:35
we could talk all evening uh about this uh very interesting facet of of AI so
19:38
uh very interesting facet of of AI so
19:38
uh very interesting facet of of AI so explainable AI is an important thing to
19:41
explainable AI is an important thing to
19:41
explainable AI is an important thing to also have in mind whether or not you can
19:46
also have in mind whether or not you can
19:46
also have in mind whether or not you can explain what the AI how the AI came up
19:49
explain what the AI how the AI came up
19:49
explain what the AI how the AI came up with the information it came up with or
19:52
with the information it came up with or
19:52
with the information it came up with or if it's actually very complicated to
19:53
if it's actually very complicated to
19:53
if it's actually very complicated to explain that and and you lose
19:55
explain that and and you lose
19:55
explain that and and you lose transparency and basically you get an
19:57
transparency and basically you get an
19:57
transparency and basically you get an answer but you don't know why
19:59
answer but you don't know why
19:59
answer but you don't know why oh wow fascinating thank you so much for
20:02
oh wow fascinating thank you so much for
20:02
oh wow fascinating thank you so much for coming to the cloud show today to talk
20:03
coming to the cloud show today to talk
20:03
coming to the cloud show today to talk to us about this Bo and this has been a
20:05
to us about this Bo and this has been a
20:05
to us about this Bo and this has been a very wonderful conversation I really
20:07
very wonderful conversation I really
20:07
very wonderful conversation I really appreciate it so I want to have you back
20:09
appreciate it so I want to have you back
20:09
appreciate it so I want to have you back sometime to talk more about a for AI for
20:12
sometime to talk more about a for AI for
20:12
sometime to talk more about a for AI for sure anyway thank you for today it is
20:15
sure anyway thank you for today it is
20:15
sure anyway thank you for today it is time to go and thank you audience for
20:17
time to go and thank you audience for
20:17
time to go and thank you audience for being on the cloud show and I'll see you
20:19
being on the cloud show and I'll see you
20:19
being on the cloud show and I'll see you again next time thank you
20:23
again next time thank you
20:23
again next time thank you [Music]