17:40:06  Live captioning by Ai-Media Live 
captioning by Ai-Media Live 
17:40:09 captioning by Ai-Media

18:04:31 SHANTHI ROBERTSON: 
After a false start,
18:04:33  I think we are now 
broadcasting to our 
18:04:36 audience. 

18:04:36 Welcome everybody once 
again stop hello and 
18:04:39 welcome to the second 
webinar in our webinar 
18:04:41 series which is brought 
to you by the ADEPT
18:04:45  project, autonomy 
disability and diverse 
18:04:46 the

18:04:48  technology project. 

18:04:49 This is a collaborative 
Australia research
18:04:58  Council linkage 
project. 
18:04:59 It is run at the 
Institute of culture and 
18:05:03 society of Western 
Sydney University. 
18:05:04 In partnership with a 
number of disability and 
18:05:07 migration service 
providers. 
18:05:08 Art organisations, and 
technology partners. 
18:05:10 My name is Chancey 
Robinson
18:05:12  and I am a chief 
investigator on the 
18:05:14 project but my main role 
tonight is to be your 
18:05:18 host for this webinar. 

18:05:20 Before we begin, I would 
like to acknowledge that 
18:05:23 our university resides 
on the Darug people
18:05:25  of the nation and I am

18:05:27  calling in

18:05:31  on their traditional 
land. 
18:05:32 They have been owners 
of the country for 
18:05:35 thousands of years and 
many of you will also be
18:05:41  participating from the 
unseeded lands of 
18:05:43 traditional owners. 

18:05:44 The project team wishes 
to acknowledge and pay 
18:05:47 our respects to the 
elders past, present and 
18:05:49 emerging of the lands, 
always, especially this 
18:05:52 week which is 
no doc week . 
18:05:54 The aim of our webinars 
series is to bring 
18:05:58 together experts

18:06:01  from the projects Corp 
research team, but also 
18:06:04 our wider networks, to 
discuss some of the 
18:06:07 critical issues we are 
all collectively 
18:06:09 grappling with 
in our work. 
18:06:11 The theme for this 
evening's webinar is 
18:06:13 crisis, technology and 
disability service 
18:06:15 delivery. 

18:06:16 I will introduce our 
expert panel now. 
18:06:18 Our first panellist is 
Lida Ghahremanlou. 
18:06:20 She is an analytics 
cloud solution architect 
18:06:23 and UI

18:06:24  architect in 
Microsoft UK. 
18:06:26 She is joining us very 
early in the morning.
18:06:29  Snow Li

18:06:30  is our next panellist.

18:06:31  She is

18:06:32  psycho social

18:06:33  project

18:06:37  social project lead. 

18:06:38 And our final expert 
is Liam Magee
18:06:41 , associate Professor 
of all.
18:06:43  To let you know how

18:06:47  this will run, we will 
start off with a panel 
18:06:50 discussion. 

18:06:51 I will ask a few key 
questions about
18:07:03  of our panel. 

18:07:04 If you have questions, 
we would welcome your 
18:07:07 contributions. 

18:07:08 Please type them in the 
Q&A function and in the 
18:07:11 second part of the 
session, I will call 
18:07:14 some of these audience 
questions and try to 
18:07:17 cover as many as we can 
for our experts to 
18:07:20 discuss. 

18:07:21 We will also have a 
couple of polls popping 
18:07:24 up for our audience. 

18:07:26 One will pop up in a 
couple of minutes and 
18:07:29 one towards the end 
of the session. 
18:07:32 It would be great if our 
audience member's could 
18:07:35 contribute their 
thoughts there. 
18:07:36 Let's get started with 
our discussion. 
18:07:38 I am going to start with 
just asking each of you 
18:07:42 a little bit more about 
the specific work each 
18:07:45 of you are doing in 
relation to technology 
18:07:48 and service delivery 
in your different 
18:07:50 organisations. 

18:07:51 I might start with Snow. 

18:07:53 I note your site has had 
to make a number of 
18:07:57 changes to how its 
services are delivered 
18:07:59 to the public for stop 
if you can start off by 
18:08:03 telling us a little more 
about your organisation, 
18:08:06 the communities and what 
your response has been 
18:08:09 to some of the issues 
that have cropped up 
18:08:12 this year? 

18:08:13 SNOW LI: We have been 
around for more than 30 
18:08:17 years. 

18:08:17 We have always been a 
key organisation 
18:08:20 providing support to 
people with disability 
18:08:22 who are aged and who are 
experiencing mental 
18:08:25 illness. 

18:08:25 Being a service provider 
in the northern Sydney 
18:08:28 region for that long 
time, we have supported 
18:08:31 lots of families through 
their life stages. 
18:08:33 This is quite unique, as 
it is for everyone else. 
18:08:37 We started with the 
bushfires and quickly 
18:08:39 followed by COVID. 

18:08:41 As an organisation, we 
have always planted our 
18:08:43 technology 
implementation plan for 
18:08:45 the organisation. 

18:08:46 We had a roadmap, what 
we would implement in 
18:08:49 technology to help 
people better access 
18:08:51 care. 

18:08:52 What happened this year, 
we had to fast track 
18:08:55 things and it really 
change the assumptions 
18:08:58 were used to have. 

18:08:59 In the care sector, 
there has or has been 
18:09:02 this assumption that 
people's disability and 
18:09:05 people who are aged 
are not tech savvy..
18:09:08  That is or is been the 
discussion every time 
18:09:11 people bring it up, 
there is or is a strong 
18:09:14 backlash of we are not 
catering for everybody 
18:09:17 if we start to engage 
with technological 
18:09:19 solutions. 

18:09:20 With COVID, moving 
everything to Telehealth 
18:09:22 helped us to

18:09:26  fast track how quickly 
people got into 
18:09:29 technology. 

18:09:29 And their assumptions 
about technology were 
18:09:31 plainly wrong compared 
to what we assumed at 
18:09:34 the beginning. 

18:09:35 Having said that, we 
do see a different 
18:09:38 adaptation process

18:09:39  of all the clients. 

18:09:40 For us, it was a slow 
transition process
18:09:43  within the organisation

18:09:47  confronting some of the 
assumptions that we had, 
18:09:50 but also helping clients 
individually to get onto 
18:09:53 technology.

18:09:54  SHANTHI ROBERTSON:

18:10:00  LINDA McKINNEY: , Can 
you tell us more about 
18:10:03 the work Microsoft is 
doing with it AI for 
18:10:07 accessibility program

18:10:10  and other services 
for disability? 
18:10:12 Particularly if you have 
any feedback from the 
18:10:14 field about how 
effective some of these 
18:10:17 have been during covert? 

18:10:19 LIDA GHAHREMANLOU:

18:10:22  Most people know 
Microsoft is 
18:10:24 traditionally a software 
company, but since 2014, 
18:10:26 Microsoft has become 
one of the bigger clad
18:10:29  providers.

18:10:30  -- Cloud providers.

18:10:34  Their role has changed 
significantly and it 
18:10:36 comes

18:10:37  within the culture of 
the company as well.
18:10:42  In terms of AI and 
automation systems,
18:10:45  we have a movement

18:10:49  in our culture

18:10:53  to provide to end users

18:10:57  that make AI accessible 
for everyone.
18:11:00  Always the culture

18:11:02  of accessibility

18:11:04  has been in a Microsoft

18:11:07  mainly around

18:11:09  the late 90s

18:11:10 ,

18:11:12  we implement 
at the culture
18:11:15  into the hiring

18:11:16  process

18:11:18  and in recent years

18:11:21  they have an initiative 
program
18:11:23 ,

18:11:24  like I

18:11:26  a

18:11:27  AI for good

18:11:28 .

18:11:29  In that

18:11:31  killer shift

18:11:32  in the company,

18:11:36  they have

18:11:40  moved into how software 
and culture and products
18:11:42  can change towards

18:11:45  making AI accessible 
for everyone.
18:11:51  It doesn't exclude any 
particular minority
18:11:54  and trying to include 
all people
18:11:57 .

18:12:07  As many big 
tech companies
18:12:09  once started,

18:12:12  they have immediately 
provided
18:12:18  through software 
technology, the company 
18:12:19 has responded with

18:12:20  a load of

18:12:26  cloud providing access 
for different customers 
18:12:28 as well as

18:12:29  minority people

18:12:31  who may not 
be able to use
18:12:36 , for example,

18:12:37  Teams

18:12:39  which has been

18:12:42  increased for people 
with particular needs.
18:12:48  In terms of vaccine 
development, Microsoft 
18:12:50 was one of the

18:12:51  earliest

18:12:54  sponsor

18:12:54  that

18:12:57  made the

18:13:00  dataset about 
COVID available
18:13:01  through the

18:13:03  open

18:13:04  dataset.

18:13:09  Many

18:13:11  use that dataset

18:13:13  to start some

18:13:14  ML

18:13:16  competition

18:13:21  to answer the questions 
clients were answering
18:13:25  asking to develop 
the caps
18:13:28  vaccine.

18:13:31  There has been major

18:13:34  help through 
the technology
18:13:36  for companies that

18:13:37  are developing

18:13:41  their own vaccines

18:13:43 .

18:13:45  There are a 
number of them
18:13:48  that are quite close 
to a result.
18:13:51  I guess that's 
the way Microsoft
18:13:54  got united.

18:13:56  There has also been 
a special request
18:14:00 , consideration,

18:14:02  full

18:14:04  employees

18:14:06  with

18:14:09  needs, who have special 
circumstances
18:14:12 , to the management 
and through
18:14:14  the different

18:14:25  advisors, they make 
sure that every employee 
18:14:27 can work at home

18:14:29  and

18:14:32  have an easy to work 
from environment at home.
18:14:37  SHANTHI ROBERTSON:

18:14:41  One of the fantastic 
things that is coming 
18:14:44 through so far as we have

18:14:46  panellists from really 
diverse organisations
18:15:01  ranging from service 
delivery to big tech. 
18:15:03 The next thing I wanted 
to ask is an issue which 
18:15:07 I think is probably 
common across all 
18:15:09 organisations in this 
space, regardless of 
18:15:14 Snowbank

18:15:18  SNOW LI:

18:15:30  We are moving a large cohort of 
people, so people who are age, 
18:15:34 people who have disabilities and 
people from different cultural 
18:15:37 backgrounds and we realised everyone 
has a different perspective. 
18:15:39 It depends on the age, cultural 
background and the disability and 
18:15:43 the capacity they have with trust 
and sharing information online
18:15:46  and communicating with 
everyone online,
18:15:51  so I think in the previous workshop 
I touched on how different cultures 
18:15:55 perceive trust different 
with privacy. 
18:15:56 They believe

18:15:59  sharing a private document with 
their peer group is completely fine
18:16:03 . 

18:16:03 When there is distrust

18:16:09  with the government, it becomes 
more problematic so when we are 
18:16:12 collecting data from people online, 
one question they often ask is
18:16:16  how are we going to use the data?

18:16:19  Will it affect my children's

18:16:21  future in 20 years?

18:16:22  It

18:16:24  involves a lot of distrust and 
depends on the social and demographic
18:16:29  profile, so for us it is to engage

18:16:32  a very individualised approach

18:16:36  when talking about trust and 
sharing information online and 
18:16:39 communicating online.

18:16:40  With the older generation 
they prefer a phone call,
18:16:51  more traditional means 
when it comes to email. 
18:16:54 Sometimes we will have a backlash 
– why do you have my email? 
18:16:59 It is a very interesting topic and 
something we have come across and 
18:17:03 have decided to take a very 
(inaudible) approach and technology 
18:17:07 with that particular intersection. 

18:17:09 Having said that,

18:17:12  COVID definitely changed the 
landscape and people's approach to 
18:17:16 technology because like it or not 
you have to rely on technology and 
18:17:21 that includes people with disability 
relying on telly health for very 
18:17:24 private sessions. 

18:17:25 Some session involves behaviour 
support when seeing a person, they 
18:17:28 are on the screen, how do you 
build trust from the screen? 
18:17:33 Then we quickly have to learn from 
that and telling our therapist
18:17:38 , ask the person with disability 
the likes and
18:17:40  dislikes

18:17:43  and build trust through 
other means. 
18:17:45 You can have a background 
of super heroes
18:17:48  if the children with disability are 
passionate about superhero so we have 
18:17:52 two use different ways to

18:17:54  build trust

18:17:55  before the service

18:17:57  delivery session so this period 
will help us come up with creative 
18:18:02 solutions in tackling trust 
and privacy issues in this
18:18:15 ?

18:18:18  space.

18:18:21  LELA KOGBARA: Microsoft 
is built on trust
18:18:23  especially because the role we have

18:18:25  is a big

18:18:28  tech company in the society 
as a cloud provider
18:18:33 .

18:18:36  It is very important and 
we have this message
18:18:39 , a slogan through the company 
that we are built on trust so
18:18:44  we with the

18:18:45  -- with the technology of AI 
because the breakthrough especially
18:18:49  for the last New Year's 
has been significant and
18:18:52  speech to text

18:18:54 , cognitive services

18:19:02  and other breakthroughs that come 
through to help the automation AI. 
18:19:06 --

18:19:06  View. 

18:19:07 The question has been asked

18:19:10  next to the extent, what about 
the transparency and privacy?
18:19:17  There have been

18:19:19  stories that I am sure the 
audience and yourself no
18:19:22  house some

18:19:25  of this privacy has been broken 
down through the services
18:19:29 , through the other major 
big tech companies because
18:19:32  there

18:19:35  has not been regulation, 
how to use that technology
18:19:38  in terms of

18:19:39  not too kind of

18:19:43  invade the privacy

18:19:45  of individuals,

18:19:47  so because of that

18:19:53 , we have next to AI for 
accessibility and all the AI 
18:19:57 breakthroughs and the tech companies 
have set their goals for, we have
18:20:02  responsible AI, which is

18:20:05  based on six major principles and

18:20:08  includes fairness

18:20:11 , reliability and safety

18:20:12 , accountability

18:20:15 , privacy

18:20:22  and transparency and in each of 
these principles we are looking into
18:20:27  a specific aspect of trust

18:20:30  and how our product can be fair

18:20:35  and not having the bias

18:20:37 , even if the data has bias

18:20:42 , how the product and software 
is able to identify that bias
18:20:46  through the data

18:20:52  and through the transparency pillar 
we have a particular software, open 
18:20:56 source that helps the scientists

18:20:58  to be able to identify

18:21:01  and understand

18:21:03  the results of an

18:21:09  ML model, an AI model that would 
identify again if there is any bias
18:21:14 , if there is something

18:21:17  we are not understanding, what 
is that and it is also about
18:21:21  looking into other accountability

18:21:23  and security

18:21:25 … We are looking at that

18:21:27 . 

18:21:28 Responsible AI

18:21:29  is not

18:21:32  a responsibility for individuals, 
it is across people
18:21:36 , government

18:21:39  and also the societies 
so it is important that
18:21:42  this concept is shared

18:21:45  not only by technology, not only by

18:21:47  individuals but also by

18:21:51  big government and 
through this practice
18:21:54 , we also have our

18:21:57  committee that make sure

18:22:00  the regulation comes 
through the software
18:22:02  and any new product 
that we make available
18:22:06  to the public follows all these 
principles I guess this way
18:22:10  the company is trying to 
stay on top of the game
18:22:14  and be as transparent 
as they can be.
18:22:19  SHANTHI ROBERTSON:

18:22:26  Responsible AI is a whole of 
society approach we are looking 
18:22:30 towards at this moment.

18:22:34  I might throw to you 
on this idea of trust. 
18:22:38 You might have a macro perspective 
on how trust might be working across 
18:22:42 different industries. 

18:22:43 LIAM MAGEE: There has been an 
interesting transition in relation 
18:22:47 to how trust is in a sense

18:22:49  fixed

18:22:52  on particular organisations.

18:22:55  Have said decade or more ago there

18:23:00  was greater concern around public 
institutions is specifically
18:23:05  governments and the use 
of data of citizens
18:23:08 , for instance and it is 
only been quite recently
18:23:19  major sort of scandals

18:23:20  where trust has been breached

18:23:23  or the perceived bond of trust 
between company and consumer has 
18:23:27 been breached

18:23:28 , so that I think Bach's 
one transition. 
18:23:31 I think this year there is in a 
sense another kind of change underway
18:23:36 ,

18:23:36  which often

18:23:40 … Trust has been spoken about

18:23:42  in relation to

18:23:45  its market or consumers

18:23:49 .

18:23:51  Marks.

18:23:54  Technology may not have

18:23:56  the kinds of

18:23:58  science

18:24:02 , technology's may not have 
the kinds of remedies
18:24:05  to particular crises and I 
think that has probably been
18:24:10  talked about a lot in environmental 
circles for a number of years
18:24:14  but in many other kinds of 
crises that are emerging now
18:24:18 , it is not always a kind of

18:24:21  situation

18:24:23  that science and technology, 
whether through companies of
18:24:27  public institutions 
can come in and solve
18:24:30  problems effectively and I think 
that particular shift is one that
18:24:35  is quite challenging

18:24:37  to process, both from within 
companies themselves but also
18:24:40  within the broader public

18:24:44  because it suggests there is a 
limit to the degree of control
18:24:53  that socially we can exercise over 
the emergence of crises and of 
18:24:58 course COVID is the most prominent 
at the moment but I think the same 
18:25:02 sort of situation around in 
a sense trust, relates to how
18:25:06  environmental crises like bushfires 
as I mentioned earlier
18:25:09  are dealt with

18:25:10 , and a final point

18:25:12  – I think it is also 
important to note as
18:25:16  Lida has done in a way 
the extent to which
18:25:24  trust has become a commodity for 
many organisations that not only do 
18:25:28 they pride themselves on a trusted 
relationship with customers. 
18:25:31 I think that concept has been around 
a long time, but it is something, 
18:25:36 almost a kind of property 
(inaudible) and just to refer to one 
18:25:40 incident that is evidence of this, 
immediately after the Cambridge
18:25:44  Analytic

18:25:50  scandal there was a moment where 
Tim Cook, the CEO of Apple 
18:25:54 criticised Facebook and contrast of 
the way Apple dealt with privacy and 
18:25:59 security for its customers compared 
with Facebook, so it clearly becomes 
18:26:03 itself a commodity to differentiate 
one company from another.

18:26:07  SHANTHI ROBERTSON:

18:26:19  Corporatisation of trust. 

18:26:20 If we think about trust, 
one-dimension is the really 
18:26:23 authentic inclusion 
of technology uses
18:26:28  and in the design of testing and 
tech products and for the work we're 
18:26:33 doing this includes users with 
disability and you all have 
18:26:37 different opinions and experiences 
on this so I might want to explore 
18:26:41 that idea a little more about 
the idea of codesign.
18:26:47  Snow, perhaps starting with you 
because you have been involved in a 
18:26:51 start-up venture that has 
codesign built into it. 
18:26:54 What you understand about codesign 
and what can it do to address some 
18:26:59 issues around technology? 

18:27:00 SNOW LI: In response to the crisis 
we have seen in the disability 
18:27:05 sector a lot of players enter the 
space being the start-up companies 
18:27:09 and they are able to adapt than 
environment quite well because they 
18:27:13 are agile with respect to products 
and product design but in this 
18:27:17 process, it is a rapid moving 
process and heavily involves them 
18:27:21 validating their idea, 
prototyping the idea and
18:27:24  testing with the users, so we 
always talk about codesign in 
18:27:27 services but I guess that a company 
because they want to be user 
18:27:32 friendly product -- build. 

18:27:34 The codesign process is well done

18:27:40  -- known so I appreciate this 
process and it could have a more 
18:27:45 personalised solution for people 
with different needs and we are now 
18:27:49 seeing more of these players 
in the disability space. 
18:27:52 At the same time, we are working 
with some of the disability start 
18:27:56 ups and being involved, I can 
quickly see how it can also be a 
18:28:01 problem if it is not trained 
and done well. 
18:28:05 One example is one of the software 
engineers was very curious about 
18:28:09 designing a solution so passionate 
about the space and wanting to go 
18:28:13 out and do some user 
testing himself. 
18:28:15 Without talking to me, so he went 
into the field and got\straightaway, 
18:28:19 so basically, in the disability 
sector we came along way. 
18:28:23 We had the process, the revolution, 
the words we use,
18:28:26  have an understanding of what the 
service is about, what it means and 
18:28:31 that is something that happens in 
the last 10 years, so continuously 
18:28:35 we evolve the words that we use in 
the service context the goods we 
18:28:40 know we want to be more person 
centred and remove the stigma from 
18:28:44 the way we speak. 

18:28:46 But when the software engineer went 
out and did the units are testing 
18:28:51 himself he was using the wrong words.

18:28:53  -- User testing. 

18:28:55 He was not from the sector. 

18:28:57 Coming back to the trust issues 
where trust is very much valued in 
18:29:02 the sector

18:29:09 . 

18:29:10 Once they interact with you, they 
find out you understand the sector, 
18:29:14 have been in the sector long enough 
and the trust was built instantly 
18:29:20 and for him using the wrong words 
talking about technology, it shuts 
18:29:23 people down, so when we are talking 
about tech technology and design, it 
18:29:27 is a methodology but at the same 
time you need to understand the 
18:29:32 sector and how to engage to get the 
most out of it otherwise you will 
18:29:37 not get anything, so I'm glad you 
brought the codesign path up and 
18:29:42 this is a very interesting area and 
I think that… I see that in the 
18:29:47 future we need to meet somewhere in 
the middle, need to be some sort of 
18:29:52 protocols or ways that can help this 
technology… The players to better 
18:29:56 engage with people with disability 
with a range of different 
18:30:00 conditions. 

18:30:01 SHANTHI ROBERTSON: I think that is a 
critical point and that is hopefully 
18:30:05 something as a community of practice 
that we can build some of those 
18:30:10 connections between tech users 
and the service sector as well. 
18:30:12 scale. 

18:30:12 That is the issue of 
trust, which often comes 
18:30:16 up is very key in 
these discussions. 
18:30:18 I might throw to you 
again, Snow, to talk 
18:30:21 about trust in your 
contacts and what that 
18:30:24 means for you 
in your role. 
18:30:26 LIAM MAGEE: I think in 
a research perspective, 
18:30:29 co-design is a powerful 
instrument. 
18:30:31 Precisely because it 
serves to break down a 
18:30:33 lot of barriers that 
often exist between 
18:30:36 researchers and

18:30:37  participants. 

18:30:38 In doing so, I think it 
can help shift certain 
18:30:41 relations of power. 

18:30:43 If you think about 
traditional research, 
18:30:45 the researcher frames 
the research, ask silly 
18:30:47 questions of the 
participants, they 
18:30:49 respond with answers and 
the researcher takes 
18:30:51 those answers away and 
analyses and comes up 
18:30:54 with areas types 
of conclusions. 
18:30:56 That general model 
doesn't disappear 
18:30:58 entirely,

18:31:01  but there are processes 
by which participants 
18:31:03 can actively get 
involved in research
18:31:05  in a sort of analogous

18:31:07  way to technology 
is an important way
18:31:10  of destabilising

18:31:13  existing structures 
of power.
18:31:17  As you have alluded to,

18:31:20  I think it's been very 
effective in the search 
18:31:24 contacts

18:31:26  context

18:31:28  to deal with products 
that aim to
18:31:32  incorporate users into 
a kind of product, 
18:31:34 whether a game or 
another type of 
18:31:37 technology. 

18:31:38 I want to note 
the co-design
18:31:40  is also referenced

18:31:42 , raises problems 
of its own.
18:31:46  There are questions 
around who gets invited 
18:31:48 to participate 
in co-design
18:31:51  workshops

18:31:54 , how accessible those 
sorts of events might be
18:31:57 . 

18:31:58 Particularly in 
the contexts
18:32:00  I have worked in, how 
co-design decisions can 
18:32:03 often be overridden 
in later phases of
18:32:05  development. 

18:32:06 They can be difficult 
to circumvent
18:32:09  because often the 
reasons for overwriting
18:32:11  are to do with

18:32:12  costs

18:32:15  and time blowouts 
and so on. 
18:32:18 Co-design has its own,

18:32:23  it can be an expensive 
process, or it came up 
18:32:26 with particular ideas 
that are difficult to 
18:32:29 realise. 

18:32:29 Finally, I think 
in the context
18:32:35  that interests us in 
our project around 
18:32:37 automation and 
artificial intelligence, 
18:32:39 it's a really difficult 
concept to know how to 
18:32:42 integrate in a 
fundamental way
18:32:45  precisely because AI 
systems are already 
18:32:47 highly complex. 

18:32:48 Even the engineers 
involved in the design 
18:32:51 are often not 
entirely sure
18:32:52  of how an AI system 
produces the results it 
18:32:55 does.

18:33:01  It opens than a lot of 
questions around 
18:33:04 advanced technology 
systems, about exactly 
18:33:05 how co-design 
will fit in. 
18:33:07 Nonetheless, there are 
certainly roles to play 
18:33:10 in areas like

18:33:11  selecting training 
data, designing the 
18:33:12 architecture of AI 
systems, evaluating 
18:33:14 results and particularly 
in the government's use.
18:33:21  Perhaps to turn to 
Lida to take this 
18:33:23 conversation forward, it 
would be interesting to 
18:33:26 hear from you about 
ways in which
18:33:28  customers and other 
kinds of stakeholders 
18:33:30 are able to get involved 
in AI design
18:33:33  in a spirit 
of co-design.
18:33:36  Lida Ghahremanlou Mac

18:33:37 \

18:33:41  LIDA GHAHREMANLOU:

18:33:45  I like the way you are 
looking at it from a 
18:33:49 research perspective.

18:33:52  That is completely the 
right conclusion that 
18:33:54 you have got. 

18:33:55 Also, I just want to 
make a point about 
18:33:58 Snow's comment about 
what care means.
18:34:01  It is something

18:34:02 ,

18:34:03  when we decided

18:34:04  a big

18:34:07  tech company design 
and started doing
18:34:09  this AI product,

18:34:10  that is the question

18:34:12  they

18:34:13  asked themselves 
later on.
18:34:15  What care means.

18:34:18  There has been

18:34:20  a transfer

18:34:22  from the design 
perspective
18:34:24  from the technology 
perspective
18:34:26  to the care perspective.

18:34:27  I can see

18:34:29  the shift has 
been significant
18:34:32  and obvious

18:34:33  that

18:34:34  now

18:34:38  the inclusive

18:34:39  design

18:34:42  coming at the centre 
of the decision and 
18:34:45 principles

18:34:45  that I discussed.

18:34:49  One of the

18:34:50  ways

18:34:52  that the company 
has shifted
18:34:54  is providing 
the employee
18:34:56  with mandatory training

18:34:59  on accessibility

18:35:02  for AI.

18:35:03  Meaning

18:35:06  we are supposed to take 
this online training
18:35:09  that we know

18:35:10  how to

18:35:16  refer and what 
terminology to use, not 
18:35:18 just the technology 
and not just AI
18:35:21  kind of way that 
we always speak.
18:35:24  That has become

18:35:25  a

18:35:25  core training

18:35:28  part for us

18:35:31  for every employee 
is essential.
18:35:34  The other thing,

18:35:38  through the 
responsibility
18:35:40  that I mentioned

18:35:43  of six principles 
that we have got,
18:35:45  we have practices

18:35:50  that seal our 
organisation,
18:35:53  making sure that 
when the software
18:35:55  has been produced

18:35:59  and is being

18:36:01  in the use of 
the end user,
18:36:03  does it follow all the 
essential principles
18:36:06  that

18:36:06  identify

18:36:08  responsible AI

18:36:12  and making sure that 
these follow those 
18:36:14 regulations.

18:36:15  Through that practice,

18:36:17  they introduced tools

18:36:21 , something like

18:36:22  interpret

18:36:24 , those

18:36:25  are open source

18:36:28  software that they keep 
improving and we are 
18:36:31 also learning 
how to use them
18:36:33  to train our

18:36:34  customers

18:36:36  and end users

18:36:39 , that when they 
use our software,
18:36:42  they make sure

18:36:43  to use these tools 
together and
18:36:45  not just

18:36:46  using, for example,

18:36:49  a learning environment

18:36:51  and just train the model

18:36:53  and see what the

18:36:54  technology

18:36:56  can do for them.

18:36:59  We have an expression

18:37:00  that

18:37:02  it is a

18:37:04 .

18:37:04 ..

18:37:08  At the end of a speech 
when we talk about the 
18:37:11 technology,

18:37:12  we say one more thing.

18:37:15  Have you thought about 
the accessibility of 
18:37:18 your products?

18:37:20  We include the 
six principles
18:37:22  that we have been 
putting in place?
18:37:25  What are your responses

18:37:27  to providing

18:37:30  the inclusive design

18:37:35  to your final product? 

18:37:37 These are the questions 
that we have been 
18:37:40 trained to ask as

18:37:41  delivery services that 
we put for our customers.
18:37:44  That is a major shift.

18:37:46  In the past, maybe 
10 years ago,
18:37:50  this was not a concern 
or maybe just a side note
18:37:54 . 

18:37:55 At the moment,

18:37:56  this is the core

18:37:58  for us.

18:38:03  Anytime we talk about 
the services and the 
18:38:06 product.

18:38:07  That

18:38:09  tells a lot of promise

18:38:14  that we are designing 
for the end user
18:38:16 ,

18:38:17  to the core of 
the conversation
18:38:19 .

18:38:20  Not just something 
as a side point.
18:38:23  I hope it answers 
the question.
18:38:30  SHANTHI ROBERTSON: 
Thank you all. 
18:38:31 These are great 
insights. 
18:38:33 I have one more question 
for the panel.
18:38:40  I would like to remind 
our audience members. 
18:38:42 If you have questions 
for our panellists or 
18:38:45 even a comment to add to 
the discussion, please 
18:38:48 to type them into the 
Q&A and we will start 
18:38:52 the interactive portion 
of addressing some of 
18:38:54 your questions and 
comments very soon. 
18:38:56 I thought we would just 
finish off with one more 
18:39:00 question for the panel. 

18:39:02 I think that's really 
about the range of 
18:39:05 organisations that 
you will come from. 
18:39:07 We have got Snow from a 
service provider that is 
18:39:10 providing service 
delivery here in Sydney 
18:39:12 to the community 
directly. 
18:39:14 We have Lida 
representing a large 
18:39:16 tech company. 

18:39:17 And we have the academic 
research perspective as 
18:39:20 well. 

18:39:20 Each of these 
organisations have 
18:39:22 different aims. 

18:39:23 As we have addressed 
in the conversation 
18:39:26 start-ups, they come 
from a framework of 
18:39:28 being fast and agile, 
attributes that can 
18:39:30 sometimes not work well 
with the slower pace 
18:39:33 provided with inclusion 
and consultation of 
18:39:35 co-design.

18:39:38  Other organisations 
have different 
18:39:40 challenges to do with 
scale and size. 
18:39:42 The university sector 
has challenges around a
18:39:49  disability and 
inclusion. 
18:39:50 With the crisis we have 
been staring down in 
18:39:53 2020, perhaps there is 
an increase in the 
18:39:56 perceived need for 
speed, for things to 
18:39:59 happen quickly in change 
and technological 
18:40:01 transformation solutions 
to be rolled out 
18:40:03 quickly. 

18:40:04 I was wondering, does 
this sometimes perhaps 
18:40:06 happen at the expense of 
some of these methods of 
18:40:10 inclusion we have talked 
about today? 
18:40:12 Snow, I will

18:40:14  throw to you first. 

18:40:16 Do you have any thoughts 
on that crisis context 
18:40:19 and the need for 
acceleration? 
18:40:21 SNOW LI: For me, it is 
important to see that 
18:40:24 there are two sides.

18:40:34  There are capacity and 
empowerment for the two 
18:40:37 sides. 

18:40:37 The service provider, 
they have more in-depth 
18:40:40 understanding of how to 
support the person, so 
18:40:43 that we do what's 
of handholding
18:40:45  support with our 
clients to transition 
18:40:47 them through

18:40:47 , to help them 
understand technology, 
18:40:50 to bring the shared 
understanding using 
18:40:52 different methods 
of technology use
18:40:56 . 

18:40:57 Understanding what sort 
of technology we are 
18:40:59 using for our clients 
and how they perceive 
18:41:02 the technology. 

18:41:03 On the other side of the 
example given earlier, 
18:41:06 the software engineer,

18:41:07  we need to consider the 
start-up plays in the 
18:41:10 disability space. 

18:41:11 Are they empowered 
enough? 
18:41:13 Do they have a capacity 
to design solutions?
18:41:15  At the end of the day,

18:41:18  is it easy for the 
user to use them,
18:41:21  they want to have 
more efficient
18:41:23  and good co-design 
workshop
18:41:24  with the

18:41:25  rental and use 
of the product.
18:41:28  You do see a need

18:41:30  to see

18:41:35  the capacity building 
empowerment for both 
18:41:37 sides of the party. 

18:41:39 How do you bridge that? 

18:41:41 What sort of FREYA: of 
concrete framework 
18:41:44 mechanism

18:41:44  can be used to guide 
both sides to better
18:41:47  reach the middle point?

18:41:52  As you pointed out, 
traditional care 
18:41:54 providers tend to have 
a very slow process, and
18:41:59  on the other side, 
the new players, the 
18:42:01 start-ups, they are very 
forward thinking.
18:42:06  They are hearing the 
start-up structures. 
18:42:08 Lots of money for the CTO

18:42:10 s

18:42:10  and money spent 
on the technology,
18:42:20  but when they talk 
about technology in the 
18:42:23 care space, it is the 
people that matter. 
18:42:26 They seem to be missing 
that point. 
18:42:28 It is not a blanket 
statement. 
18:42:30 I am hoping to see the 
middle point is bringing 
18:42:34 together a group 
mechanism, and efficient 
18:42:36 mechanism for 
the project. 
18:42:37 SHANTHI ROBERTSON: 
Thanks. 
18:42:38 Lida, what about you? 

18:42:40 Do you have any thought 
about crisis times?
18:42:43  LIDA GHAHREMANLOU: 
Absolutely.
18:42:45  I total agree with 
all the comments
18:42:48 .

18:42:50  We realise

18:42:51  that there is a gap

18:42:53  between the technology

18:42:56  and the delivery 
services that we provide
18:42:59 .

18:43:02  One of the greatest 
practices that we are 
18:43:05 trying to do

18:43:06  in the past 
few years, is
18:43:08  through the

18:43:11  programs

18:43:12 ,

18:43:13  we worked

18:43:18  closely for six months 
with start-ups. 
18:43:20 We bring them in, we 
help them to build
18:43:23  and we are here for 
what they are looking for
18:43:27  in terms of their 
service delivery. 
18:43:29 We learn from them 
how to design
18:43:31 ,

18:43:32  what the design 
needs to be
18:43:35  and what needs 
to be included
18:43:39  and excluded. 

18:43:40 We bring our expertise 
with the
18:43:42  training and 
the knowledge
18:43:43  that the start-ups bring

18:43:45  and work

18:43:48  with them. 

18:43:49 We help them 
to build this
18:43:51  in a prototype

18:43:52  kind of

18:43:54  product.

18:43:57  That practice for me,

18:44:00  I join

18:44:02  always AI for good

18:44:04  programs, practices

18:44:06 .

18:44:10  That practice for 
me is very valuable.
18:44:13  I get in touch 
with the actual
18:44:16  start-ups

18:44:18  and companies

18:44:21  who are doing 
the real design
18:44:23 .

18:44:25  We're just providing 
a platform for them.
18:44:28  This is a great shift

18:44:31 , let's say for

18:44:33  a big

18:44:35  thing, always

18:44:37  thinking to produce more

18:44:42  and just take the 
market for themselves.
18:44:45  This is a

18:44:49  way to make them stop, 
think and listen to 
18:44:53 them. 

18:44:53 Try to give them

18:44:55 ...

18:44:55  The

18:44:58  thing they think they 
need to design.
18:45:04  They have the data from

18:45:08  people with low vision.

18:45:11  There is a tracker 
they are trying to
18:45:16  design, it gives 
them the
18:45:21  location of longitude 
and latitude of that 
18:45:23 person. 

18:45:24 They are looking to be 
the prediction model
18:45:27  that can help them

18:45:28 , for example when the 
person is trying to turn
18:45:32 , doesn't match 
with the location
18:45:34 .

18:45:35  Is it a safe 
place to turn
18:45:38  two

18:45:38 ?

18:45:42 Liam, I thought I might throw back 
to you now about the idea of 
18:45:47 codesign. 

18:45:48 Do you have anything further you 
wanted to reflect on there?
18:45:57 For me,

18:45:58  that is a very valuable

18:46:00  working environment together

18:46:03  because I could immediately respond

18:46:06  with technology

18:46:08  and the services that we have

18:46:11  but I have somebody sitting there

18:46:13  working with

18:46:15  low vision users and they are 
telling me something that I
18:46:19  do not know through 
the data and I guess
18:46:22  these sorts of practices

18:46:23 , I think I want to conclude that 
these sorts of practices and 
18:46:28 workshops that we run through the 
company give us the opportunity to 
18:46:32 bring a codesign concept

18:46:33  to the core of our technology 
development and I believe this is 
18:46:37 the vision for Microsoft and 
we are moving towards that
18:46:41  and I'm very happy to be able to use

18:46:44  the background and the 
technology that we know
18:46:47  for these very valuable reasons.

18:46:49  SHANTHI ROBERTSON:

18:46:56  We are keeping such a close eye on 
the work you are doing, Lida, to see 
18:47:01 how this plays out. 

18:47:03 Liam, I might get your thoughts, I 
know that we often grappling with the
18:47:16  tension between Capel 
and ethical processes
18:47:18 , so do you have anything

18:47:20  to add there?

18:47:23  LIAM MAGEE:

18:47:28  I was involved in the development 
of the agile software if you like and
18:47:33  what is interesting is reflecting

18:47:35  on ways of inclusion and codesign, 
participatory design is in certain
18:47:39  respects the agile movement

18:47:42  which made it popular or popularise 
this idea of moving fast
18:47:45  and doing things at speed

18:47:51 , was in terms of its detail really 
focused on slowing down that it 
18:47:56 really brought the four listening

18:47:58  as Lida has mentioned, listening 
to customers or users of products, 
18:48:02 building in testing early on and 
a process called peer programming
18:48:08  that would have people sitting 
side-by-side at one keyboard which 
18:48:12 in theory would halve the rate 
at which software was developed
18:48:16  and yet, because this would

18:48:17  precisely kind of lead

18:48:24  to something a more reflexive 
process of software production, it 
18:48:28 would be agile and more responsive, 
so I think sometimes we are forced 
18:48:32 and I know the kind of questions we 
develop for this session put forward 
18:48:37 this kind of dichotomy if you like 
between care and patience versus 
18:48:41 moving fast

18:48:44 , but in certain respects, at least 
in technology, it is often the case 
18:48:49 that agile means precisely taking 
time and taking time to be 
18:48:53 inclusive, to involve people 
more systemically
18:48:55  physically in the disability space 
so I will conclude on that point.
18:48:59  SHANTHI ROBERTSON:

18:49:04  Thanks. 

18:49:05 I'm not getting any questions 
and answers from the chat. 
18:49:08 Don't be shy. 

18:49:10 We still have a few minutes left in 
this webinar and we have SME people
18:49:15  -- smart people on the panel but no 
questions to big or small, so feel 
18:49:21 free to type something in but I 
might take the liberty to keep the 
18:49:26 conversation going. 

18:49:27 I thought that scope quite macro

18:49:29  and think big picture

18:49:35  stuff,

18:49:37  so

18:49:42  for 2020, what has been the broad 
role of technology in relation to 
18:49:46 the crisis we face this year? 

18:49:49 I might start with you 
this time, Liam. 
18:49:51 LIAM MAGEE: I think

18:49:55  to go back to a point are made 
earlier, one of the interesting 
18:49:59 points relates to this trust 
and the sense in which
18:50:05 , I think in the very first 
instance, certainly not for people 
18:50:09 who are medical practitioners or 
involved in medical research but I 
18:50:13 think the public is probably 
surprised there was not a cure all 
18:50:17 this did not resolve itself, 
COVID did not resolve itself
18:50:21  much more quickly and we were 
forced into these progressive 
18:50:24 lockdowns,

18:50:27  the governments had hugely varying 
responses to the crisis that there 
18:50:31 was no global consensus about ways 
to kind of deal with it and even
18:50:36  expertise

18:50:37  seemed to be split often although 
clearly some kind of consensus has 
18:50:41 emerged, and I think that

18:50:43 , in a certain sense, is profoundly 
shocking in the sense that there is 
18:50:48 no standard way in which

18:50:50  crises can be analysed

18:50:54 , rationalised, developed into a 
programme of operation that is then 
18:50:58 carried through in ways 
that are universally
18:51:03  recognised and of course that has 
always been a myth but I think the 
18:51:08 extent to which countries and even 
within countries communities have 
18:51:11 been deeply divided by

18:51:13  how best to respond to crises, 
how even to interpret it
18:51:22 , points to a deep fundamental 
questioning around how we look@is, 
18:51:26 technology and how we understand its 
role in determining what courses of 
18:51:30 action we should take and also 
perhaps some surprise that 
18:51:33 technology has not been… Although

18:51:38  as Lida mentioned it has been 
critical in the work of companies 
18:51:42 like Microsoft, critical in 
gathering data sets and assisting 
18:51:46 drug and pharmaceutical companies in 
looking at vaccines, things like 
18:51:49 contact tracing have been very

18:51:51  hit and miss, I think, interns of 
the efficacy from country to country
18:51:55 . 

18:51:56 Conversely, the important 
role of technology that
18:51:58  has been played

18:52:08  out has been around mundane 
functions around communication, 
18:52:10 entertainment and facilitating 
remote work and session such as this.
18:52:14  SHANTHI ROBERTSON:

18:52:15  Snow, how about you, did 
you have any big picture
18:52:21  reflections on technology 
and the 2020 pandemic? 
18:52:23 SNOW LI:

18:52:24  How adaptable we are, so

18:52:26  helping lots of clients transition

18:52:30 . 

18:52:31 I do not know what Zoom 
is, I cannot set it up.
18:52:35  But to the stage where 
you see everyone
18:52:41  interacting on the screen, we take 
people on virtual tours of things 
18:52:45 and there work looking lessons,

18:52:47  the connection through technology, 
something if I talk about that, 
18:52:50 before COVID,, "No, there 
is no way you can connect
18:52:55  through a computer screen

18:52:57 ,"

18:52:59  to how quickly you can 
see people doing exercise
18:53:03 , to cooking in the kitchen, 
everyone sharing the same
18:53:06  recipes

18:53:07 .

18:53:09  "

18:53:12 This is something we made. 

18:53:14 It looks the same."

18:53:15  It also confided my belief that 
only close contact is important.
18:53:19  People can quickly connect 
with each other and adapt
18:53:23  to their environment

18:53:27 , old or not. 

18:53:29 The craving for connection can still 
easily be facilitated through 
18:53:33 technology and how important it is 
to see them smiling in front of 
18:53:37 cameras and seeing the impact 
straightaway through the computer 
18:53:41 screen, so I think that it is again

18:53:43  so many assumptions were made in 
the sector about technology use and 
18:53:47 how engaged with technology 
and sessions, tele-health
18:53:50  and those barriers quickly got

18:53:55  dissolved two or three months into 
COVID, so again, for me, the biggest 
18:54:00 lesson is how adaptable we are in 
how we can better use live technology
18:54:05  for good things. 

18:54:06 SHANTHI ROBERTSON: Thanks

18:54:08 . 

18:54:09 That is a good thought. 

18:54:11 We have been pretty adaptable. 

18:54:13 Lida, what about you? 

18:54:15 You are about to embark on 
a second lockdown in the UK.
18:54:19  Do you have any reflections on 
the broad role of technology
18:54:23  in what we have all been 
through this year? 
18:54:26 LIDA GHAHREMANLOU: I guess

18:54:28  I

18:54:29  agree with Liam

18:54:34 's,. 

18:54:35 I agree with Snow we got 
adaptable very fast.
18:54:38  I guess

18:54:39  people like me who work with this

18:54:41  high-technology

18:54:43 ,

18:54:46  which is like immediately the 
companies provided us with a lot of
18:54:51  facilities like if you needed a 
second monitor, they could send it 
18:54:55 to us

18:54:57 . 

18:54:58 The features they added 
to Zoom or to Teams
18:55:01  and we also only

18:55:03  got set up and we all 
had an office at home
18:55:07  and that is a great part

18:55:09  – I agree on that

18:55:11  – but I think we kind of missed 
out on some of the comments
18:55:16 , some of the other aspects 
of working from home
18:55:20  and dealing with crises and

18:55:22  you only notice this

18:55:23  when you put it

18:55:24  into this perspective

18:55:26  with other people.

18:55:28  Yesterday

18:55:30  in our tech holder, 
which we used to get
18:55:33  (inaudible)

18:55:36  and I find that screen 
time is so overwhelming.
18:55:39  You have to be

18:55:49  in front of a screen from 8 AM to 
5 or 6 PM and often working hours 
18:55:54 continue. 

18:55:55 You don't get a proper lunch breaks 
of these are all the disadvantages 
18:56:00 of this adaptability that we have. 

18:56:02 But we had a colleague with a 
disability on the call and she was 
18:56:07 talking about her own experience 
working from home
18:56:10  and technology, the fact she has 
a similar environment like us
18:56:14 , and I was just

18:56:17 … She mentioned because 
of her condition
18:56:20 , she could not leave the house

18:56:25  since March because she needed 
volunteers to help her to leave the 
18:56:29 house and because of 
the extensive time
18:56:32 , like three months of lockdown

18:56:34  in early spring

18:56:35  and it just crossed my mind

18:56:38  how we notice

18:56:44  these minority people and how our 
technology that we claim to have
18:56:50  a lot of breakthroughs

18:56:54  like speech to text and 
text-to-speech and vision, all the 
18:56:58 services with vision (inaudible) 
the NHS tracking
18:57:00  kind of software

18:57:03  being broadly used in the UK

18:57:08 , has been used to register 
and I just felt that
18:57:13  it is a great picture

18:57:14  out there but

18:57:17  there are a lot of

18:57:22  small holes that we do not see 
unless somebody puts it into 
18:57:26 perspective that, "You forgot 
about this and that,"
18:57:29  and I guess you cannot 
blame the technology
18:57:34  in a broader perspective because 
everyone hundred years
18:57:37  human societies faces

18:57:39  these is

18:57:45  -- disasters.

18:57:48  At a college in London, 
they published
18:57:50  a report that

18:57:53  we will deal with this 
crisis in 18 months
18:57:56  and I remember when 
I read that report
18:57:58  I thought it was a joke.

18:58:02  We are going to get through 
it in six months. 
18:58:06 Our society would not…

18:58:08  But we are nearly 12 months now and 
we still haven't found a vaccine
18:58:18  and the governments are not putting 
the right measures in place and still 
18:58:22 dealing with it, so I guess a lot 
of learning through this crisis
18:58:27  for individuals

18:58:27 , for organisations 
and for government
18:58:30  as well and the best way is

18:58:35  just to put it into perspective

18:58:45 . 

18:58:45 I would say that is the main thing 
– not only looking at our own 
18:58:51 perspective, looking at different 
perspectives and then you get the 
18:58:54 right measure to see how 
we deal with crises
18:58:57  from a global perspective. 

18:58:59 SHANTHI ROBERTSON: A very important 
point to finish on. 
18:59:02 We have learned to lock there are 
gaps to fill and we have to thinking 
18:59:07 about it from the perspective other 
people living in other contexts and 
18:59:12 circumstances. 

18:59:12 We do need to wrap this up. 

18:59:15 I want to thank all of 
you for participating. 
18:59:18 Thank you so much, our three 
incredible panellists, for
18:59:21  those insights and

18:59:49  the interesting conversation, thank 
you to the Auslan interpreters for 
18:59:52 your support. 

18:59:53 Before you go, I would like to 
remind everyone we have one more 
18:59:58 webinar in the series this week on 
Tuesday, 17 November from 2 PM to 3 
19:00:03 PM, as slightly different topic next 
week, engaging our audiences online, 
19:00:07 reputation, diversity and disability 
in the art and creative industries so 
19:00:11 watch out for info about that. 

19:00:13 We hope you can join us and thanks 
so much, everyone, for being here. 
19:00:18 It was a really fruitful discussion. 

19:00:21 LIDA GHAHREMANLOU: Thank you. 

19:00:22 LIAM MAGEE: Thank you. 

19:00:24 Live captioning by Ai-Media