Guest lectures

17:02:03 Okay, I think we can start. I'm gonna just briefly introduce the speaker. So, hello everyone.
17:02:11 And welcome to our second. Guest lecture organized by the Open Research Network. So today we are very happy to have Dr.
17:02:20 Sharon Black with us and she's from the University of East Anglia. She will talk to us today about the value and applicability of using the qualitative research methods when doing research and media accessibility with participants.
17:02:36 So it will be particularly interesting, to anyone working with interviews and working with end users. We very excited to have you and thank you so much for accepting the invitation.
17:02:48 And the floor is yours, Sharon.
17:02:52 Thank you, much, Alicia, and thank you to you and the whole of the open research network for inviting me.
17:02:58 For having me here today. At yeah to give this talk i'm really honored and i hope you'll find it useful I am still, it's still very much, it's, it's a bit of a work in progress.
17:03:15 This really This idea really came out of a comment at a conference from a colleague. Who sort of commented that this was at the EST conference in Oslo.
17:03:33 Trying to remember dates and places and I presented some research there some qualitative research that I'm going to present today.
17:03:42 As part of the really great question and answer session and discussions. Somebody sort of said, well, you know, these are just opinions.
17:03:53 They're just opinions and, That really, I really, this is one of the reasons why I really enjoy engaging in conversation and debate and discussion at conferences.
17:04:03 Because that's where sometimes you get challenges or where people really You know, provoke. Debate and get you really thinking, you know, of course my reaction was.
17:04:15 Well, it's not just opinions, isn't it? And, but they got me thinking about why, why is it not just opinions, why is it important?
17:04:23 And then I started thinking about sort of the. Prevailing prevailing methods and approaches within our field.
17:04:32 It really got me thinking about that. And what I think really, where the directions that we need to develop.
17:04:41 Our research and media accessibility. So I'll talk about that not today, but I just really wanted to.
17:04:48 You know, start with that sort of starting point that was really inspiration for, for this and made me realize that I needed to.
17:04:57 I probably needed to argue for this. I couldn't just take it as a given. So yes, talking about the value and applicability of using.
17:05:05 Quality of methods and media accessibility research with participants. So we're hoping my arrows are going to work, yes.
17:05:13 So, a little bit of background. You know, as I'm sure you all know, media accessibility studies really began.
17:05:26 As part of audio visual translation as a subfield of translation studies. It's really, I suppose academically it was really born as the area of AVT.
17:05:40 AVT scholars did embrace accessibility as part of their research failed. But in recent years, media accessibility has really grown, steadily grown to become an independent field in its own right.
17:05:55 According to Greco and Jan Kovska and I'd very much agree with this.
17:06:02 Generally I've noticed that AVT and MA are presented alongside each other. So they sit alongside each other and many Headings and many and many academic texts, AVT and MA.
17:06:15 And but is also growing into an area of accessibility studies as Greco and Jankowski have pointed out and it's really becoming a driver for major social change.
17:06:29 And you know as an independent. Failed in its own right. You know, and It's really.
17:06:39 But that's talking, that's talking about today, you know, coming up to today, but.
17:06:44 In fact, audiences have always been pivotal, pivotal. They've always been key and essential in research.
17:06:51 On translation studies and right up to today talking about media accessibility today.
17:07:00 And so therefore, along with this rise in media accessibility research, there's also this, this increase, this burgeoning reception.
17:07:11 Research both within audio visual translation and media accessibility. As noted by the Giovanni and Gambier and Anna Jankhowska, for example.
17:07:24 However, I would argue that, thus far, reception studies in media, in media accessibility, have and and an AVT research have focused.
17:07:36 Mostly on collecting quantitative data and employing cognitive experimental. Approaches conducting psycho linguistic research.
17:07:50 This is, I'm also involved in this kind of research, so I, I promise I'm not biased, but I've note noted that you know there's been a real focus I think.
17:08:00 On these approaches on quantitative approaches using a battery of questionnaires, tests, you know, comprehension tests for example, many other kinds of tests.
17:08:14 Electrothermal activity, eye tracking research. Measuring heart rate for example And I have listed a few studies within, you know, of this type within media accessibility, but certainly there are other studies, this is just to to mention a few and certainly some of these studies do you employ qualitative methods as well, you know, for example, interviews and focus groups.
17:08:43 But these are, these tend to be secondary, very much secondary and you know, complementary, but also not really focusing on the.
17:08:54 The methodology, the qualitative methodology and approach. Are used that tends to be sort of tagged on.
17:09:02 At the end. Yeah, these, these studies are extremely valuable, of course. You know, they test, for example, users reactions to different linguistic and technical criteria.
17:09:14 Investigating cognitive processes behind. you know, the reception of, media accessibility.
17:09:24 Surfaces. Certainly valuable insights. Have been gained and are being gained and will continue to begin. I would argue that there's very much a need to employ more qualitative approaches.
17:09:39 And maybe accessibility. Research. To to gain more in depth understandings a richer picture of users views and experiences as well.
17:09:51 So actually I've sort of begun to answer that question already. Yes, why is there a need to employ more qualitative?
17:09:58 Approaches. Well, So to talk, I think you should talk a little bit about what qualitative research is.
17:10:10 And what is what is its value in general but also focusing I mean there are many forms of qualitative research you know conduct such as focusing on texts.
17:10:23 Case studies but here as you'll have noticed from the title I'm focusing on qualitative research with participants and in particular what we have been calling reception research.
17:10:37 But I'm maybe you've thought about this too, but reception is quite a passive.
17:10:42 Concept I would argue it's a passive activity but more about that and just a moment So, well, qualitative.
17:10:51 Research involves collecting and analyzing non-numerical data. As I've mentioned, it could be from text, using video, audio.
17:11:00 For example, to understand people's views. Perspectives, experiences. And qualitative research really has a rich tradition in the study of human social behavior.
17:11:15 Since the late nineteenth century, researchers have been involved in studying social behavior and the cultures of humankind.
17:11:24 And even since the nineteenth century, researchers have really perceived the limitations. And using in quantification in in in terms of explaining the human, phenomena and behavior that they have encountered.
17:11:46 So really, academic, all academic disciplines concerned with human and social behavior such as sociology, psychology, anthropology really do make extensive use of qualitative research.
17:12:02 Quality of research methods really enable us to, to understand how people experience the world. And can be used to, to gather in-depth insights into your problem or to generate new ideas for research.
17:12:17 Among other things.
17:12:20 So what are the traditional qualitative methods? I'm not here to sort of get, you know, give you, I could talk.
17:12:29 For an hour and many more hours about. Qualitative methods and so on but just to refresh our memories What are the traditional qualitative methods?
17:12:38 So they are observations, which involves recording what you've seen. Heard or encountered using detailed feedback.
17:12:48 And that involves spending time spending quite a bit of time with the research participants. That you're interested in observing, getting experience of really spending time with them, building up that rich experience as a researcher and knowledge.
17:13:03 And insight. Also interviews. Which can be semi structured unstructured or structured which involves asking people questions.
17:13:16 It says one and one conversations here. I mean, generally, yes, interviews are one on one. I would argue also, for example, I've conducted interviews with children and I decided to conduct those interviews with pairs because I felt that it would be Maybe less daunting for the children being interviewed.
17:13:36 But also easier for me to control in pairs rather than in larger groups. Of kids or maybe interacting with each other so that's just an example hi for example it might just be not always be one on one And also focus groups, you can conduct focus groups.
17:13:55 With your, as the name suggests, with your participants in groups. And that really helps generate discussions.
17:14:02 That that can be really fruitful in terms of generating discussions among your participants. And surveys yes surveys are kind of a hybrid form of data collection aren't they?
17:14:16 So because questionnaires, I would say. Mainly predominantly quantitative. However, you can also have open-ended questions in your surveys which can be a really effective way of collecting qualitative data.
17:14:32 But I would argue, you know, it's tempting to pick the methods that you're familiar with.
17:14:38 And perhaps there's been a tendency. For those of us embarking on some qualitative research.
17:14:44 To stick to those. Traditional methods, but why not, Why not think about more creative research methods?
17:14:54 Not just I mean, we'll talk about the benefits in just a moment, but really The point I wanted to emphasize here is that sometimes those traditional methods are not adequate and maybe they're not enough to fully address your research topic.
17:15:09 The complex kinds of research topics that you that you might want to address. You know if you, so these methods can be useful for as I've said addressing complex research questions.
17:15:24 Particularly in relation to new trends or aspects of social life. Does that sound familiar to you, maybe as media accessibility researchers?
17:15:33 At complex research. New trends, new aspects of social life. Where traditional methods may not be sufficient to answer those questions.
17:15:46 These methods can also be useful if you research is focused on problems solving. Which may also resonate with you.
17:15:55 If your topic is new or emerging, we certainly have plenty of new and emerging areas within media accessibility.
17:16:03 Or if you're research project has a specific, in challenging, changing, or responding to dominant assumptions in society.
17:16:12 Again, I feel this, you know, may resonate with you as, You know, scholars with an interest in media accessibility.
17:16:21 So the methods I've got here, on the slide are, first of all, arts based research.
17:16:29 So briefly, arts based research. Can focus on it focuses on using artistic processes. And various forms.
17:16:39 Britain, visual, spoken, performance. You know using video film filmmaking dance you know various creation of artistic forms.
17:16:51 As a main way of understanding and exploring experiences of a topic. Both by the researcher and the participants.
17:16:59 So these methods can be really effective in exploring people's experiences, sensitive topics. Exploring feelings and emotions.
17:17:12 People who speak different languages again very relevant and people with communication difficulties for example so I think very, relevant to us.
17:17:24 Research using technology and what I mean here is interacting with your participants via online communities or studying your participants via online communities.
17:17:35 Using, for example, digital tools to create things. Blog posts could be through social media, and virtual interviews and focus groups.
17:17:46 Which I think some of us have experienced during the pandemic and have become more experienced. And and running these types of sessions.
17:17:57 Transformative research can be very useful for if you're aiming to challenge or change what might be seen as a dominant position or a status quo in society to challenge or change part of your participants lives.
17:18:14 You know, some sort of transformation. These transformative research can include participatory research, action research and community based research.
17:18:26 Which, I'll show you some examples today. Also mixed methods research. I'm sure many of you, perhaps all of you are familiar with mixed methods research, but this is really this idea is a combination.
17:18:44 It's often sort of as using a combination of quantitative and qualitative methods. Such as traditionally the questionnaire and the interview.
17:18:53 But there's actually more to this concept and it can, involve combining more than one qualitative method.
17:19:01 I hope I've provided inspiration. By mentioning quite a few here, you know, in terms of creative methods that can be mixed.
17:19:09 And the mixed methods can be useful to consider when you're seeking to explore questions or topics that are too complex.
17:19:16 For one method alone to address.
17:19:22 So, talking a little bit more about mixed methods. So and also in addition to addressing complex topics.
17:19:35 Mixed methods can also be valuable in the exploratory stages of a research project. So you can use.
17:19:43 The research I'm gonna present to you today was exploratory in a way and I find so many.
17:19:48 So many different things were brought up by, you know, really valuable and important insights from participants that could all go on to be their own, you know, through all the further explored.
17:20:02 So and the exploratory stages to verify research questions or refine research questions. Do you assist with conceptualization?
17:20:13 Or to generate a hypothesis. This is in terms of mixed methods. So if you're thinking of going on, say to an experimental study.
17:20:24 I can also help talking about experimental. Researching the qualitative methodologies can also help with identifying the correct variables.
17:20:34 You know, at sometimes if there's not much existing research in your area. It's, it's often worth thinking about whether you are actually identifying the correct variables to be measured.
17:20:48 Maybe qualitative expiratory research would be useful first. Because you might find that you're not targeting the most appropriate.
17:20:56 Factors. If you don't do that. Quality of work can also be valuable for.
17:21:03 Interpreting, qualifying or illuminating quantitative research findings. So.
17:21:13 Yeah, it can be complementary or it can contradict. You also might find their areas where your qualitative data contradicts your quantitative findings.
17:21:24 Also for triangulation. And supporting the validation. So for example where 3 or more methods are used.
17:21:33 And the results are compared for similarity. So lots of different uses there if you're thinking of doing or you're currently embarking on mixed methods.
17:21:43 Therefore, so what have said so far I would, assert at this point that qualitative methods can be more effective than quantitative approaches in explaining complex phenomena.
17:21:57 I think I have. Already mentioned this this idea of the complexity and highlighted the, the, you know, complexity.
17:22:09 And as such, qualitative methods are valuable additions to your Methodist methodological armory.
17:22:14 They're not just something we can tag on at the end. As a compliment without really putting too much thought into the methodology, but really they are valuable additions to our methodology.
17:22:30 So, but. You know, methods are not always, you know, it's important to select the best methods for your research questions.
17:22:40 And not just use a method for the sake of it. So what research questions are best answered using a qualitative approach.
17:22:49 So, it's important to remember that, first and foremost, unlike quantitative research, Inquiry conducted in the qualitative tradition.
17:23:02 Seeks to answer the question, what? As opposed to the question how often. So qualitative methods are designed to reveal what is going on by describing and interpreting phenomena as well as it says here.
17:23:16 They don't attempt to measure how often. How often something happens. But rather, it's all about describing and interpreting.
17:23:27 Phenomenon.
17:23:33 And, research conducted using qualitative methods is normally done with an intent to preserve the inherent complexities in human behavior.
17:23:44 As opposed to assuming a reductive view of the you know, of the topic. In order to kind or measure the occurrence of phenomena.
17:23:58 So it's really about, preserving the, preserving the inherent complexities rather than reducing.
17:24:07 Or taking a reductive approach.
17:24:10 So that means that there's not always easy answers. The You know, the an experience I faced when I presented my qualitative research.
17:24:22 At EST. Consequence of this was that I wasn't able to present the audience with easy takeaways with easy answers in terms of what you know in terms of what you know the voices of the users and the experiences that I'd had and That's maybe not something that audiences are always, always, accustomed to being faced with in terms of their thoughts, you know,
17:24:55 we'll talk, talk more about that in a moment, but yes. It's very different in a way than presenting.
17:25:01 You know key quantitative findings. Really? In that sense. Quality of research normally takes an inductive approach.
17:25:10 So, often moving from. Observation to hypothesis rather than hypothesis testing or deduction.
17:25:19 So often At the moment I am deep in analysis of my qualitative data and I'm I've started there really.
17:25:31 And then I'm going to, you know, from there. Go to the research and compare the what I find for my data with the research, you know, rather than.
17:25:42 The other way around, saying an eye tracking study, I would start with the existing research. And work from there taking a deductive approach.
17:25:51 So it's it's very different approach. Why am I arguing then for, Oh, sorry.
17:25:59 Yes, I think there's maybe. Do I have Yeah, I have, I have, sorry, I've altered the order of my slides.
17:26:10 So. Yeah, I what I wanted to add here was some. I wanted to counter some of the misconceptions about qualitative research.
17:26:19 So I've mentioned at the beginning that a comment I encountered at a conference was well it's just opinions simply just opinions and other conceptions are that qualitative researches of little value.
17:26:35 This is obviously not my view, but, this is, an existing view. That there's little value in qualitative research because we cannot generalize from them.
17:26:47 So in case you're not aware, generalization. Refers to the extent that findings can be applied.
17:26:56 To other people. or other times or settings than the, so other people to the population more generally.
17:27:04 Rather than just the the people in taking part in the study. And a common criticism of qualitative research is that the results of a study are rarely, if ever, generalizable to a larger population because the sample.
17:27:20 GROUPs are small. And qualitative research. And they're not chosen randomly.
17:27:25 Hi, I would argue that such, you know, these criticisms feel to recognize the distinctive of qualitative research.
17:27:36 So whereas in quantitative research the intent is to secure a large random sample. That is representative of the general population.
17:27:47 You know, focusing on generalizations and thereby allowing for statistical inference of results.
17:27:53 That aim to be applicable across the entire population. On the other hand, in qualitative research, Yeah, generalizability is based on the assumption.
17:28:05 That it's valuable to begin to understand similar situations or people. So understanding people that are similar in one way that have shared characteristics, they possess relative relevant characteristics.
17:28:20 For the question being considered. So. Qualitative researchers often use what I don't know if you've heard of this, but propulsive sampling.
17:28:32 So I know I'm sort of. You know, presenting a lot of concepts here, but it's really, Just an overview, you know, I'm happy to talk more at the end of the talk, you know, discuss these concepts more if you'd like.
17:28:47 And tell you more but propose of sampling is where participants are selected deliberate at but deliberately to test a particular question.
17:28:59 So it's not about identifying a random group but really group, a group that possess relevant characteristics as it says here.
17:29:07 Re to the question. So the idea is to have a rich interpretation then that deepens our understanding.
17:29:18 It's about richness and in-depth understandings. Of phenomena within that group.
17:29:27 That's been selected. As a result of their relevant characteristics for the question.
17:29:36 Another common misconception is that qualitative research How not really claim reliability or validity. So briefly in quantitative research, reliability, is the extent to which different observers can make the same observations or collect the same data.
17:30:01 About the objective study.
17:30:07 So, but in. In qualitative research, we have this concept. I don't know if you've heard of it before, but it's concept of.
17:30:17 Trustworthiness.
17:30:20 And trust one way to demonstrate trustworthiness in qualitative research is to present detailed evidence. In the form of quotations from interviews and field notes.
17:30:30 Along with a thick textual descriptions. Of episodes. Advance settings in observational research, for example.
17:30:42 To be trustworthy, qualitative analysis should also be audible. Making it possible to retrace the steps.
17:30:52 The leading to a certain interpretation or a certain theory. To check that no alternatives were left unexamined and that no research or biases.
17:31:03 Had any avoidable influence on the results.
17:31:08 In terms of validity. In, in the sciences and the natural sciences or in in experimental studies, findings are validated by replication.
17:31:21 And if a second investigator cannot replicate? The findings. Then it's found that, you know, the original results are rejected as flawed and invalid.
17:31:33 So, that's, a, very briefly.
17:31:42 However, in using qualitative research, for example, triangulation can be used. Combining multiple views, approaches and methods.
17:31:54 To obtain a more accurate interpretation of the phenomena. Thereby creating an analysis of greater depth or greater richness.
17:32:04 So. This is really.
17:32:09 Oh, way in which I can go more in depth, you know, into more depth on this, but that's really a way that we can increase validity and qualitative research.
17:32:24 So that's just I've talked about qualitative research in general really there and maybe I hope that as I was presenting those points that you felt that you could relate them to media accessibility, you know, and perhaps to your own research interests.
17:32:42 But, looking more specifically at media accessibility. Research. Why am I arguing as it asks here is that more qualitative research is needed.
17:32:54 And media accessibility.
17:33:01 So, well.
17:33:05 One of the reasons is that what we're, what we're seeing now in media accessibility research We're seeing, we've seen a change.
17:33:14 You know, over time in recent years. So until recently audiovisual translation studies and media accessibility studies focused on modes of consumption.
17:33:28 On audiences passively consuming audio visual. Content that's been made accessible. But really in recent years, changes in modes of consumption and technology.
17:33:44 In recent years well it sounds slightly more than a decade but i think it's probably longer now you know times move quickly but Approximately.
17:33:55 There's been an unforeseen surge in agency interactivity on the part of audiences as DIY on the part of audiences as D.
17:34:02 Giovanni and Gambia have pointed out. And Yeah, so. There's more agency and creativity.
17:34:12 And personalization in the hands of users as it says in the in the title here. Today media content can be created.
17:34:21 Shared. Modified and customized by by users. Therefore the empowerment of end users has really As did Giovanni and Gambia have said.
17:34:32 And stir the very notion of audience of what an audience is. And that, you know, there's not an increasing difficulty in marking any boundary really between media and the producers and consumers.
17:34:51 Of translated and accessible, media.
17:34:56 Some are different terms used to talk about this, producers, I think, Elena that Giovanni used and, mentioned in, in her work on this.
17:35:12 Jan Maria Greco. Talks about the, the, I hope I'm pronouncing this right.
17:35:19 I couldn't produce it as poetic, but I think it's poetic. Who is poetic?
17:35:23 I think I need Jan Maria to confirm that for me but we can see it here on the slide.
17:35:29 So the poetic model of agency. And even though my pronunciation isn't great, I'm certainly really interested in this.
17:35:37 This concept, this model. Because Jan Maria has explained that this is a model whereby all stakeholders, including users, Have a voice and are fully participating active agents and co-creators in in media accessibility process processes.
17:35:57 There's also, we're also experiencing greater personalization. Aren't we, you know, greater personalization in, media?
17:36:08 That media products that that we that we interact with that we use. And with all of this comes increasing complexity.
17:36:18 You know, again, this word complexity comes up.
17:36:22 So we, we are no longer passive audiences, but rather active users. Of digital content and products that and we may you know users who make informed decisions you know he should be able to make informed decisions about how they interact with and you know interact with accessibility services and with accessible media.
17:36:50 Content. To really benefit fully from that from those experiences and those interactions.
17:36:58 Also, qualitative research can help to give voice to marginalized people, marginalized groups.
17:37:06 And, So yeah, you may be familiar with some of the concepts presented here on the slide.
17:37:17 That really provide a theoretical framework. So, you know, it's relevant to, us, you know, with those of us who have an interest in media accessibility research with participants.
17:37:35 So for example, the social model of disability. According which according to the social model of disability you prefer your many of you will be familiar with this model.
17:37:45 But briefly according to the social model of disability. Disability is understood as a form of social inclusion, sorry, exclusion.
17:37:57 I'm so used to talking about inclusion, but we also have to talk about exclusion. It's important to do so.
17:38:02 So disability understate is a form of social exclusion imposed on disabled people by a society which oppresses and disables them.
17:38:12 And this, This. It comes from the article by Tom Shakespeare the social model of disability.
17:38:23 And also the fundamental principles of disability.
17:38:30 On this model, the social model of disability is linked to What Jan Maria Greco calls the social model of accessible in his article towards a pedagogy of accessibility published in 2019.
17:38:47 And so this So this involves the right to active participation.
17:38:54 You know, it's the right to active participation is really embedded in the social model of accessibility with all stakeholders having a voice and a role in the co construction of solutions to access problems.
17:39:10 Also social justice. The concept of social justice is really central to media accessibility. Due to its demand for a fairer and more equitable world with respect to high wealth.
17:39:24 Opportunities and privileges, including knowledge. Are distributed within society. This this code comes from a PhD, PhD thesis by Zoe MERS.
17:39:39 On training professional re speakers to subtitle live events in the UK. Aative model for access and inclusion.
17:39:49 So when users are not, able to make full use of media content, G in part to a lack of accessibility.
17:39:58 This is an example of social injustice.
17:40:09 The social model of accessibility. And that I've just presented in the previous slide is also reflective of the 3 shifts of John Maria Greco's 3 shifts.
17:40:21 Which I am sure many of you are are familiar with. So, I'm not go to you into depth.
17:40:28 On those. I'm happy to. Expand further to discuss further at the end of the talk but briefly, Greg goes 3 shifts that he has identified.
17:40:41 And occurring in the various research areas, booking, focusing on accessibility are number one for from particularist accounts to universalist a kind of access.
17:40:51 Number 2 from Maker Center to user centered approaches and 3 from reactive to proactive approaches. I'm also aware, I think, John Maria actually also gave a talk for this network which I watched the recording and I think he he talks about the 3 shifts.
17:41:10 Possibly in his talk. But I wanted to focus today on the second shift.
17:41:16 Because, the second shift. Really to the evolution and understanding that users are bearers of valuable knowledge.
17:41:25 For the investigation of accessibility processes and phenomena. And These, if you want to know, if you want to read more, these, this information comes from Greco's, 2,018 article, The Nature of Accessibility Studies.
17:41:44 So given that according to Greco's second shift, users or bearers of valuable knowledge for the investigation of accessibility processes and phenomena.
17:41:57 Then they are experts in their experiences. And, research approaches should take this into account that users are experts on their experiences.
17:42:10 And that they, they contribute valuable knowledge.
17:42:13 Not just opinions. But yes, valuable knowledge on. On meeting accessibility.
17:42:23 I'm not the only, I'm aware I'm not the only researcher in media accessibility who has called for more you know qualitative research and media accessibility Pablo Romero Fresco has also called for more qualitative research.
17:42:39 And as he says, placing the focus back on the individual. Also Romero Fresco's, work.
17:42:50 Using an engagement based approach. Which takes the user as a reference. Coating romo fresco there Also, Romero Fresco, Dangerfield and their work on accessibility is a conversation.
17:43:07 Have, made a proposal for creative media access for alternative, media access.
17:43:17 To give you a few examples, give you a few examples of. But these are not by no means the only researchers who have been conducting research using qualitative approaches.
17:43:29 But some recent examples are Kit Dangerfields. Essay film that she presented at Media for All.
17:43:41 On access as a conversation. And participatory approaches in media accessibility. Another notable example is I just quoted Zoemers, PhD thesis.
17:43:57 Mers used action research, action research methodology. And that really facilitated close collaboration with users and providers in, research.
17:44:08 They used focus groups, observations and, you know, to really gain in depth understandings.
17:44:20 And of the, you know, the views of different groups including deaf, deaf and hard of hearing people and non-native speakers of English and Re speakers.
17:44:29 Also recent work presented at the EST Conference by Nina Riviere's and Sabia Hanoon.
17:44:39 I know you, I hope I'm pronouncing that correctly. A touch of museum to scale the collaborative development of accessible art experiences So this work is in collaboration with a variety of stakeholders again, art experts, researchers, access providers.
17:44:56 You know the number one aim of this research as they have they stated was to involve users with varying abilities in access creation and in academic exploration.
17:45:06 So here we can see the, you know, bringing, bringing bringing users together with you know, other stakeholders and, and you know, the complexity of these, of these projects.
17:45:20 So I'm aware of time and I'm really glad I focused on on these different aspects because I don't want to focus too much on my own research but I'll just being aware of time I'll give you a whistle stop to of a recent study that I conducted I've conducted that I'm still it's still a work in progress in terms of writing up the research.
17:45:40 And I presented this at the EST conference. So. Yeah, but in case you didn't have the opportunity to attend so this research is, the presentation was entitled, A Difficult Art Getting It Right.
17:45:56 A thematic analysis of users views and experiences of audio description in the UK.
17:46:05 This research was conducted with 14 participants aged 21 to 60 years with a range of different disabilities.
17:46:13 9 of the participants were blind or partially sighted. But they were also participants. who are you with a range of different disabilities.
17:46:26 Including dyslexia, and cognitive disabilities. As I've already already discussed.
17:46:35 I was, you know, a really a real key concept in this research. Was viewing users as bearers.
17:46:42 Of valuable knowledge. And the so I conducted interview research. And this was during the pandemic. So it was by telephone or Zoom.
17:46:54 And the interviews really ranged in duration from between 30 min to 2 h. So some really in depth conversations.
17:47:02 Second class. And so I used thematic analysis. To identify themes across the data set. So semantic analysis as it says here, it involves developing themes or patterns of meaning across the data set that addresses a research question.
17:47:26 And I'm sorry, I'm just going back for a second because really going back to my research question at the moment as I mentioned earlier is very exploratory and open.
17:47:37 I really was looking at users views and experiences of audio description. This is the element of the research that It was ready.
17:47:47 Ready enough to present. It's ready enough to present to you to you now, but it's still a work in progress.
17:47:52 And I identified 6 themes. From the data, from the interview data. So the things were better provisioned for contemporary media access.
17:48:05 More diversity and representation and audio description. A description tailored to young adult audiences.
17:48:14 Sharing audio described viewing experiences with family and friends. Provision of more, subtle, minimalistic, a day.
17:48:24 These are the words, these are the voices of the participants themselves. Integrating audio description into the filmmaking process.
17:48:32 So from from what the participants told me, these were the 6 themes that I identified. In terms of what they were highlighting.
17:48:39 In terms of what was important to them.
17:48:44 So in terms of better provision, for contemporary media access, the same. Participants.
17:48:54 Express that the existing provision was very much appreciated, but that there was a clear need to improve access.
17:49:01 Particularly access to older classic shows, live programming, websites. Social media, for example, video and demand.
17:49:11 Interviewees highlighted a lack of audio description, for example on graphs, in, during the pandemic.
17:49:18 You know, which is really key information for, citizens. And, interviews also highlighted that it wasn't always clear that audio description is even, it's not always clear that it's even available or easy to turn.
17:49:35 Hardy, which is the same pseudonym for a 22 year old university graduate taking part in the study said.
17:49:42 I think it adds to my experience. Whereas other people might not feel that way, but I like learning about the race of characters on TV.
17:49:53 So that links to. That actually links to the theme on diversity. How do you talk about, diversity?
17:50:02 And you know that it's important to her. And she likes learning about the race of characters on TV.
17:50:12 Entering is like how do you also, highlighted the importance of tailoring audio description to young audiences.
17:50:21 And the importance of young adult young adult voices doing audio description.
17:50:29 How do you says when I watch young a young adult program they have a voice which is more suited to that type of content which is quite nice.
17:50:37 I think Hadi is talking about Netflix here. And she highlighted she really appreciated that there was a voice more suited to that type of content for young audiences.
17:50:47 And felt that there was real thought going into it.
17:50:53 William, certain and again for a 30 year old interview he works for a sight loss charity. Said no matter how sensitive.
17:51:03 It still takes an element away from that moment in the harm movie. Oh, this is again, talking about audio description and how he would prefer the audio description to be more subtle or minimalistic.
17:51:20 So yes, he says, no matter how sensitive, it still takes an element away from that moment in the horror movie.
17:51:26 Or that you have to dim the sound of the explosions and the action to be able to describe what's going on.
17:51:35 Howia also talks about romantic. So this is today with the idea of the subtle, minimalistic.
17:51:41 Audio description. How do you says, I find it funny to have this juxtaposition of this crazy thing happening on screen and a really bored voice.
17:51:51 I think the same thing with romantic stuff. Always makes me laugh. It makes people awkward to hear audio description that is disinterested on what's happening on the screen.
17:52:01 I could talk more about this but I'm aware of time. So. Yeah, William says it can really great on me.
17:52:10 Sometimes when there's too much video description, you want it to be really subtle. I think it's a difficult art getting it right.
17:52:16 When I've watched a film that does get it right, it feels fantastic. You just get really immersed in the moment.
17:52:25 Martina, as, he was aged 51 at the time of the interview and works for a communications regulator.
17:52:35 And works for a communications regulator. And in relation, this quote is in relation to the theme of the interview and works for a communications regulator.
17:52:40 And in relation, this quote is in relation to the theme of family sharing experience with family. Martina says, if we had an accessible app.
17:52:43 And you can access the soundtrack with audio description in the app, in the app. You can watch with family, with your family.
17:52:51 Martina really felt that her family didn't wouldn't want to hear the audio description, which maybe is goes against what?
17:52:58 Way as researchers might think. And proposes would like an accessible app that would allow her to access the ID separately, but also watching with the family.
17:53:11 William. Says, he actually, you know, Interviews actually identified really identified this concept that we as researchers are familiar with that of integrating the audio description into the process into the creation of the the the media the product the film or the TV program This was identified by Bye, by the interviewees.
17:53:40 For example, Williams says, if the audio description could have some input from the artistic creators of the film, it gives us much more of a similar experience to everyone else.
17:53:52 So just to sum up, I've covered the themes just in a very genital way here.
17:53:57 So very much a whistle tops, a whistle stopter. And yeah, so just very briefly to say that it was a small sample size in this study, wherever.
17:54:13 And it did, it really did enable me to gain a rich picture of users views and experiences gained through those in depth discussions.
17:54:18 And as I said earlier, there were many insights, insights gained on areas for a very for very necessary future research.
17:54:27 Qualitative research. So thank you very much. I'm aware I've talked for a long time, so I hope that there's still, I can certainly stick around for questions.
17:54:36 And I hope others can. And yeah, I certainly welcome any questions.
17:54:42 Thank you. Sharon. No, don't apologize. I think you made a very solid case for qualitative research.
17:54:50 We actually have 2 questions in the chat already so I suggest that we start with this and then anybody else can just like raise their hand.
17:54:59 So we have a question from Kat that says is universal design included in or linked with the social model of accessibility.
17:55:08 Yes, great question. And yes, this is a really important question as well. And I was actually writing about this recently in an article related.
17:55:19 I was writing about the social model of accessibility. Would you mind if I go back to, well, maybe it's too much hassle to go back to the slide.
17:55:27 No, let me just let me just try and so that yes, the social, the, you know, the concept of universal design.
17:55:36 Is important isn't it it's key because it's that idea that when we rather than focusing on small subsets of people and we rather than focusing on small subsets of people and users and in terms of say if we're designing training and research.
17:55:54 That really it involves everyone accessibility is for all users. But how does that work? How does that work?
17:56:04 In terms of the social model access of accessibility. Well, simply, simply not simply, but it's it's important.
17:56:12 We can't. It's a difficult one. So we are including everyone in the concept of a universal design.
17:56:23 Whenever at the same time it's important to remember. That people have different fears different challenges have different experiences different needs and requirements, you know, as we as we already know and we cannot ignore.
17:56:38 We cannot ignore that. Or brush over the issues of social justice. And the social model of disability. So.
17:56:48 I feel, you know, in my work I have I find that they are. Compliment that they work together they've worked together for me so far you know so for example in the article I was the research I've been writing up.
17:57:02 So it was designed with the universal design. Concept at its core and it meant that we were able to bring together users with many different users with different disabilities.
17:57:17 And different experiences. Who really found value actually in. In talking to each other and learning from each other.
17:57:27 You know, and so my research still, you know, acknowledged very much acknowledged. The social model of disability and the importance of that, you know, at the same time.
17:57:38 I hope that makes sense.
17:57:42 So I see there's another question in the chat.
17:57:46 Oh yes, well that's yes that's a quite a difficult question. So what if the findings based on qualitative methods contradict the quantum of methods.
17:57:56 Yeah, that's difficult, isn't it? I mean, there's no conclusive answer.
17:58:02 Thank you.
17:58:06 Thank you.
17:57:56 I think that's also sorry. I think there's a follow-up question from Kat. Also, Kat, you can just, because instead which groups were involved
17:58:09 Okay, so which groups were involved in my research?
17:58:15 Yeah, Kat, please go ahead.
17:58:18 So the groups.
17:58:16 Yes, yes, like, Was it any people who experience some sort of visual issues or audit auditory or issues or like people with autism or different forms of physical.
17:58:41 Disabilities or developmental disabilities involved.
17:58:47 Yes. All of those. All of those, all of those that you pointed out. So, yes, we, in this.
17:58:56 And in this research. The research brought together. What we did was we created training. On as part of the research on using digital accessibility tools.
17:59:08 And focusing on training as well. Tools and training and the research brought together a group of people who participated in the training.
17:59:19 And we also, some of whom participated in the interviews and we also had other participants. You know additional participants taking part in the interviews.
17:59:31 And yes, in the, we have participants with, several participants with autism. Dyslexia.
17:59:41 And participants who were deaf and hard of hearing blind and partially sighted. And also participants with physical disabilities.
17:59:52 And other I'm trying to remember there was, as I don't have it on a slide from memory, but yes, essentially participants with multiple disabilities and a range of different disabilities who work together in the training and reported that they really, you know, seems overall really, you know, enjoyed that experience.
18:00:20 And it also was reflected in the the research. Here i mean today today in this the findings that i've that i've presented today and the quotes are from participants who are users of audio description.
18:00:36 And who are all blind or partially sighted. However, the broader research involved participants with different, disabilities.
18:00:47 Apologies if that wasn't particularly clear. And I think I was probably rushing at the end when I was.
18:00:52 No, no problem.
18:00:52 Hello, presenting my particular project. I hope that answers your question.
18:00:56 Yeah. Sounds like really interesting that so many different categories of people were included, to get this overall picture.
18:01:08 Of you know What's what their needs are like from their standpoint. So yes, thank you for, for the answer and yeah, it was.
18:01:25 It was. Very clear. Thank you.
18:01:29 I'm really glad it was clear. I'm glad it was clear. And I'm really glad that, You find it interesting.
18:01:37 I'm happy to carry on discussions. You know, I'm happy. I'm sure, Lisa, I can pass on my contact details and so on.
18:01:48 Yeah.
18:01:45 So feel free to get in touch. You know, if you want to carry on the discussion or have any questions.
18:01:54 I think Sharon, if you still have time, could we just go back to the question by ever because it looks quite interesting.
18:02:00 So she asked in the chat.
18:02:00 Yes, and now. I'll send an email. Asking for contact details and maybe sources.
18:02:08 So, yeah, thank you.
18:02:12 Yeah, please do. Please do. Thanks.
18:02:15 Okay, so the question is, what if some findings based on qualitative methods contradict those based on quantitative methods?
18:02:22 Which should be considered more significant and which should we try to do more research on?
18:02:29 I mean, that's a really interesting question, as you say. So yeah, thanks for that question.
18:02:33 I think it really depends on your research project. So it depends on your findings. And your methods. And your.
18:02:44 Your interests, your aims as a researcher. Really? Sorry I can't be more specific, but it is it is a very general question.
18:02:54 So it really would depend. But it's I'd say it's important not to be.
18:02:58 Worried or concerned? It's not a bad thing. You know, if you're your findings contradict each other.
18:03:06 This is part of, you know, research is messy and should be messy. I mean, of course it's nice if everything is consistent.
18:03:13 But one of the things, this sort of puts me in mind of that sort of contradiction and messiness.
18:03:21 I think that's one of the things that really we were discussing when I presented my research at EST was that.
18:03:29 It's messy and contradictory and so what do we do? What do we take away?
18:03:32 What do we do now? Basically, and I think that's one of the challenges of quality, you know, doing qualitative research with participants.
18:03:42 And learning more about their views and experiences is that it's complex. And nuanced and can be contradictory.
18:03:51 You know, but it's also, you know, rich and detailed. It really enhances our understanding.
18:03:58 And, I think we have, I think we really need. As part of maturing as a field to conduct more qualitative research to learn more.
18:04:09 About our you know, about users, views and experiences and perspectives. But I wouldn't, yeah.
18:04:18 It really, coming back to the question, it really depends. But I wouldn't be afraid to, They conduct to feel, I think there's a feeling that within the field maybe there's been a sense that.
18:04:34 Quantitative methods are more rigorous or more valid. As I'm using a fancy machine and and I actually do that I mean in my tracking research I think we really shouldn't forget about the value of qualitative research.
18:04:47 As Cat says in the And the comments that shows the complexity of humanity. It certainly does. Yeah, fully agree with that.
18:05:02 Yeah.
18:04:59 Grace, I agree with everything you just said. I also use a lot of qualitative methods in life in my project.
18:05:08 Great.
18:05:07 So yeah. Okay, then are there any other final questions before we finish? Oh yes.
18:05:19 Hi, Sarah. Hello.
18:05:17 Hi, Sharon. Thank you for your talk. I was very happy, to hear that you were talking about positive methods, because I used a lot of that on my own research, so that was really nice to see it getting and the attention that it deserves.
18:05:35 My question it's about you mentioned trustworthiness in terms of providing all the data or like the transcripts of your interviews and focus groups.
18:05:47 To I suppose prove that you're you're not being you you're not being too biased but At the same time, should we not as researchers because we're doing positive research?
18:05:58 Should we not also be really upfront and acknowledge our own biases which come from our own experiences.
18:06:06 So for example like me and you know I'm a like 30 year old white woman and who's not blind so that's obviously going to influence then how I conduct my research so you think that we should influence then how I conduct my research so you think that we showed as researchers being more, conduct my research.
18:06:27 Yes, I fully agree with you. I was trying to go through my slides and find the relevant slide, but then I don't want to take time away.
18:06:36 But yes, this concept of trustworthiness. Is it here? No, or somewhere here.
18:06:44 I think it is. Anyway, yes, there it is. So, yes, so this This idea of trustworthiness, I think I fully agree with you.
18:06:56 I sometimes when I, some of the sources that I've been, I've read. About qualitative research.
18:07:03 10 to emphasize objectivity and view the researcher as an objective. 2, nearly in a way?
18:07:14 You know a tool that can for conducting the research. You know, and that,
18:06:23 So you think that we showed as researchers being more upfront to better on biases when we do quality Yeah, but, so the findings reveal themselves to the to the researcher without the researcher actually doing anything.
18:07:25 Yeah. Okay.
18:07:32 Exactly, yeah.
18:07:25 Exactly, exactly. The researcher is completely is just a tool by which the truth is revealed. And yeah, so that is, yeah.
18:07:37 I agree with you. I think that we should acknowledge our own. biases and I think it's important to do so.
18:07:45 I think that talking about, you know, discussing trustworthiness, I think it has more to do with quality and rigor in research and transparency.
18:07:53 You know, I think it's important to be rigorous and transparent. And to follow signed ethical procedures and research.
18:08:08 Yeah.
18:08:05 You know, this is good research practice really. And this is, and if we demonstrate that we are ethical, rigorous researchers that follow signed.
18:08:16 Procedures. Then hopefully people reading our research will will trust, will trust that will trust us and see that our research is trustworthy.
18:08:29 However at the same time Yes, I think it is important to and to not shy away from recognizing the the biases that we all have.
18:08:38 I think that's part of a mature approach, you know, developed approach to qualitative research.
18:08:44 And we shouldn't be afraid because again that's one of those claims that's made against qualitative research.
18:08:49 It's biased as if that were a bad thing as if that were a to discount qualitative methods.
18:08:57 But yes, rigor and, and trust this idea of trustworthiness I feel is important.
18:09:07 Yeah, anyway. I hope that makes sense.
18:09:09 For sure, no, I agree, I agree. Thank you.
18:09:12 No problem.
18:09:18 Okay, I think I'll just, that should we reveal any disabilities we might have to increase trustworthiness.
18:09:28 Wow, what a question. Yes, I, you know, that's such an interesting question. I think of course that's up to the researcher.
18:09:35 Of course that's the researchers to whether they wish to do that. I certainly think that, You know that could be hugely valuable.
18:09:45 Hugely valuable for a research project. You know, and that's a really, really good example of high of what a researcher can bring to the research.
18:10:01 You know we're talking about objectivity and the researcher. Being completely objective and so on but actually what value you know the huge value that a researcher can bring.
18:10:12 To the research in terms of their own experiences, their own lived experiences. And, you know, so certainly that can be a huge value and I think there, you know, there are methods, there qualitative methods.
18:10:25 That you can use to, it's not my area of expertise, but I can certainly look into it.
18:10:32 Methods that involve very much. You know the researchers views and experiences involved in the research as well.
18:10:42 Okay, certainly.
18:10:41 Hmm. Okay, great. Thank you so much, Severe. Thank you for being with us.
18:10:49 And yeah, I think I'm afraid we might have to finish there.
18:10:52 Yeah, thank you for thanks again for inviting me.
18:10:55 Yeah, thank you. And, bye everyone, see you next month for our recent seminar.

0:1:37.627 --> 0:1:47.967 Alicja Zajdel Gian Maria Greco is a senior research fellow at the University of Macerata in Italy.

0:1:48.727 --> 0:2:17.177 Alicja Zajdel He holds a pH. D in philosophy and a pH. D in translation studies, and his research focuses on theoretical and applied issues in access and accessibility studies, translation studies, disability studies and human rights. He complements his research expertise with an extensive experience as an accessibility consultant for public institutions and private organizations regarding policies, live events, museums, and cultural heritage.

0:2:17.877 --> 0:2:27.367 Alicja Zajdel He's also a member of the Scientific Board of the Journal of Audiovisual Translation and a board member of the American Way of my association, among others.

0:2:28.227 --> 0:2:35.597 Alicja Zajdel And today he is joining us to talk to us about artificial intelligence in media accessibility. And this is the first.

0:2:36.937 --> 0:2:41.287 Alicja Zajdel Guest speaker lecture in our lecture series for which we are very grateful.

0:2:41.957 --> 0:2:46.197 Alicja Zajdel So without further ado, I will give the floor over to January Egri.

0:2:51.957 --> 0:3:9.997 gianmaria.greco@unimc.it I'm sorry, my mic I was was off. Ohh, good evening everyone. Thank you so much. To Alicia and Kim and open reserve. Ohh the center and the network to for the invitation.

0:3:11.297 --> 0:3:13.127 gianmaria.greco@unimc.it It's so nice to see.

0:3:13.947 --> 0:3:17.567 gianmaria.greco@unimc.it Uh, so many names I can recognize. And so.

0:3:18.557 --> 0:3:24.377 gianmaria.greco@unimc.it And many others. I can't. Can you? You can hear me, right? I am assuming that so.

0:3:25.77 --> 0:3:25.477 gianmaria.greco@unimc.it Uh.

0:3:25.547 --> 0:3:26.757 Alicja Zajdel Yes, always get.

0:3:27.987 --> 0:3:28.517 gianmaria.greco@unimc.it Yet.

0:3:30.247 --> 0:3:30.657 gianmaria.greco@unimc.it You.

0:3:30.637 --> 0:3:32.447 Alicja Zajdel And we can see the slide as well, perfect.

0:3:32.557 --> 0:3:37.237 gianmaria.greco@unimc.it That's even better. So, uh, I'll try to speak.

0:3:39.327 --> 0:3:51.837 gianmaria.greco@unimc.it I can't really promise it, but let's see. My intention is to speak half an hour, not more, because I really wish to use this chance to discuss.

0:3:53.227 --> 0:3:56.427 gianmaria.greco@unimc.it The topic with you and.

0:3:58.177 --> 0:4:0.927 gianmaria.greco@unimc.it I'm pleased to discuss this or or.

0:4:1.787 --> 0:4:5.477 gianmaria.greco@unimc.it Or would you within this context because?

0:4:6.117 --> 0:4:16.927 gianmaria.greco@unimc.it I from what they got the the goal of this network, the network puts together gathers together.

0:4:17.387 --> 0:4:36.837 gianmaria.greco@unimc.it Ohm doctoral researchers and postdoctoral researchers and part of my, or actually the core of the argument, which I will the the point I will make towards the end. I think it's very important.

0:4:37.317 --> 0:4:43.177 gianmaria.greco@unimc.it Uh, for this specific target, so for you.

0:4:44.367 --> 0:4:52.437 gianmaria.greco@unimc.it Now I have slightly changed the title. As you can see or first let me just start my.

0:4:53.477 --> 0:4:54.7 gianmaria.greco@unimc.it Timer.

0:4:56.697 --> 0:4:57.267 gianmaria.greco@unimc.it Ohh.

0:4:59.337 --> 0:5:11.637 gianmaria.greco@unimc.it Their initial title was the need for an agenda will tell you later. If I don't, please remind me why I've changed it. It's just a matter of.

0:5:13.547 --> 0:5:19.317 gianmaria.greco@unimc.it Trying to avoid a kind of misunderstanding, OK.

0:5:20.327 --> 0:5:24.947 gianmaria.greco@unimc.it To for a reader or a listener, in this case, to put.

0:5:25.657 --> 0:5:36.747 gianmaria.greco@unimc.it A things into water. I I'll be saying that it's not that are not my intention. So this is more or less the plan of the day.

0:5:37.827 --> 0:5:52.847 gianmaria.greco@unimc.it We will have a quick look at some of things from the recent past in media accessibility. Then we will talk about The Imitation Game.

0:5:53.727 --> 0:6:17.357 gianmaria.greco@unimc.it Ohm, which will lead us to talk a bit about the our role of human beings as makers expert in users. Then we will move on to discuss the refer to the human machine continuum and the human machine intensity.

0:6:19.877 --> 0:6:50.807 gianmaria.greco@unimc.it And then from there we will move to see some of the attitudes and roles towards AI attitudes towards AI roles within the discussion and research. The debate on AI. Then we will go back to media access. This should be Part 2, non part one and we will conclude with this idea of being a watchful of the need to be watchful.

0:6:52.7 --> 0:7:0.117 gianmaria.greco@unimc.it Now I had a IA was asked by Kim and Alicia to.

0:7:1.387 --> 0:7:2.37 gianmaria.greco@unimc.it Ohh.

0:7:3.387 --> 0:7:7.337 gianmaria.greco@unimc.it To send you a list of of or readings.

0:7:9.717 --> 0:7:30.127 gianmaria.greco@unimc.it The first one and the only one I ask you to read, is an article by Alan Turing. We will discuss it in the next slide, plus some other things that you may or you may not know. They are not really important. I will summarize the call.

0:7:31.7 --> 0:7:56.287 gianmaria.greco@unimc.it Ohh, but let me start with the story that I will not comment on the story I will right now I will just tell you this quick story and then as apparently nothing to do and then ohh we will come to this towards the end. So the story has Alice and Bob. Alice is a.

0:7:57.897 --> 0:8:12.897 gianmaria.greco@unimc.it Had at one point in her life tripped over quantum mechanics and she spent quite some times studying quantum mechanics than one point in her life. She tripped over architecture and she became an architect.

0:8:13.397 --> 0:8:17.147 gianmaria.greco@unimc.it Ohh and Ohh again.

0:8:17.917 --> 0:8:30.977 gianmaria.greco@unimc.it After a while, she started to join the dots and see in her mind she's in her mind that Ohh. Some of the things that I remember I studied from quantum mechanics.

0:8:31.57 --> 0:9:0.507 gianmaria.greco@unimc.it So I really use it for or could be useful in architecture, not as an analogy. OK, I'm not talking about that, I'm talking about technically, so you need to know about this. So then one day she meets Bob, Bob asked, asked her what you're up to. You have been up to and she does. Ohh. I'd been thinking about quantum mechanics and architecture and all nice. Maybe give me something to.

0:9:1.187 --> 0:9:18.547 gianmaria.greco@unimc.it So we can discuss it further maybe so she signs A3 page or she has struct structure takes out to 3 pages from my 500 book on Punta mechanics, or send it to Bob say Ohh here are some of the hints I've been working on and.

0:9:18.627 --> 0:9:19.67 gianmaria.greco@unimc.it So.

0:9:20.487 --> 0:9:42.247 gianmaria.greco@unimc.it Ohh, something like a week later. Bobby is delivering a presentation of these stories adeyola. Now that's the story and it's not the point of my I'll try to highlight has nothing to do with the ethics. So ethics in terms of, you know, being correct. In this case there's something else that interests me in this story.

0:9:44.87 --> 0:9:52.177 gianmaria.greco@unimc.it Ohh so let. Let's have a quick look at the recent past of media accessibility media accessibility.

0:9:54.807 --> 0:9:59.857 gianmaria.greco@unimc.it And has been a discussing internally.

0:10:1.37 --> 0:10:10.567 gianmaria.greco@unimc.it A for a long time, it was a focused on some aspects or some practices were.

0:10:10.967 --> 0:10:31.417 gianmaria.greco@unimc.it Uh. Uh. Consider to be the the the standard. OK became usual, but then over the past few years, my media accessibility when I say media accessibility, I'm talking about, obviously the people I've been discussing.

0:10:31.577 --> 0:10:39.927 gianmaria.greco@unimc.it Oh, oh. Questioning in a different forms, the previous assumptions.

0:10:41.367 --> 0:11:4.497 gianmaria.greco@unimc.it Ohh so for example we've been seeing for some or the idea that accessibility does not. It's not something that concerns us exclusively some people because they are not normal according to some sort of dominant concept of normality decided by.

0:11:4.997 --> 0:11:6.367 gianmaria.greco@unimc.it A.

0:11:7.787 --> 0:11:18.317 gianmaria.greco@unimc.it A dominant group indeed. I by something that concerns everyone because everyone is different. So the because the versity is part of being human.

0:11:19.687 --> 0:11:21.957 gianmaria.greco@unimc.it Ohh and so there are.

0:11:22.727 --> 0:11:27.397 gianmaria.greco@unimc.it As many forms of accessibility as there are people, basically in a way.

0:11:29.767 --> 0:11:31.427 gianmaria.greco@unimc.it Ohh then.

0:11:34.537 --> 0:11:38.127 gianmaria.greco@unimc.it Media accessibility has been somehow questioning.

0:11:38.547 --> 0:11:41.957 gianmaria.greco@unimc.it Uh, the something that was.

0:11:42.647 --> 0:11:47.347 gianmaria.greco@unimc.it Uh, right to stand up in a way. Ohh for some time.

0:11:49.77 --> 0:11:56.877 gianmaria.greco@unimc.it That was the maker center. Then expert centered approaches and the shift towards a.

0:11:57.897 --> 0:12:0.507 gianmaria.greco@unimc.it You. I see it as all the user centred approaches.

0:12:0.867 --> 0:12:7.317 gianmaria.greco@unimc.it But now looking at users as bearers of knowledge as well as active creators.

0:12:8.207 --> 0:12:11.977 gianmaria.greco@unimc.it With our top semantic love OK of their own life.

0:12:13.197 --> 0:12:13.857 gianmaria.greco@unimc.it Ohh.

0:12:15.517 --> 0:12:31.87 gianmaria.greco@unimc.it And ohh I don't. I've been other changes so one other change goes towards the direction or proactive approaches, so accessibility for a long time was I.

0:12:31.167 --> 0:12:36.297 gianmaria.greco@unimc.it Ohh was really conceived as something to be done.

0:12:36.387 --> 0:12:56.387 gianmaria.greco@unimc.it A. After that creation of some product now of ServiceNow, access is increasingly becoming what least there is this idea no that access should become an issue since the very beginning of the process.

0:12:57.27 --> 0:13:15.517 gianmaria.greco@unimc.it Ohh it's not that now everything is sold is sorted out and I access is always there since the beginning. But there is a change in there awareness and this is different awareness at different awareness for each of these.

0:13:15.607 --> 0:13:17.857 gianmaria.greco@unimc.it A 3 points.

0:13:24.97 --> 0:13:25.817 gianmaria.greco@unimc.it So far for the.

0:13:26.767 --> 0:13:35.227 gianmaria.greco@unimc.it For the recent past of media accessibility, just to give you.

0:13:35.987 --> 0:13:38.897 gianmaria.greco@unimc.it And idea? I'm not saying that.

0:13:39.37 --> 0:13:39.607 gianmaria.greco@unimc.it So.

0:13:42.907 --> 0:13:46.97 gianmaria.greco@unimc.it Universalist or user center or practical approaches?

0:13:46.537 --> 0:13:59.827 gianmaria.greco@unimc.it Ohh are something recent that would that that were not present before, but there were not the the usual the traditional that the standard.

0:14:0.397 --> 0:14:3.897 gianmaria.greco@unimc.it A way of doing or even thinking about accessibility.

0:14:4.257 --> 0:14:17.427 gianmaria.greco@unimc.it A. Obviously they have been around for a while. Just think about Gregory Fraser's or MA thesis. Usually that thesis 1975, I think.

0:14:18.27 --> 0:14:26.477 gianmaria.greco@unimc.it Ohh is considered to be the first scholarly ohh work on audio description.

0:14:27.777 --> 0:14:28.887 gianmaria.greco@unimc.it What are the description?

0:14:30.57 --> 0:14:35.607 gianmaria.greco@unimc.it You know long for longer than that, but that's the first scholarly work. At least that's that's what.

0:14:35.687 --> 0:14:40.277 gianmaria.greco@unimc.it Ohh scholars who looked into this like Anna Annamjankowska.

0:14:40.867 --> 0:14:41.197 gianmaria.greco@unimc.it Uh.

0:14:45.917 --> 0:14:46.437 gianmaria.greco@unimc.it Palos.

0:14:48.207 --> 0:14:55.277 gianmaria.greco@unimc.it And in that thesis, if we look at that thesis or there is something very, very interesting.

0:14:55.927 --> 0:15:6.557 gianmaria.greco@unimc.it Ohh, the moment he he in the 3rd it's the third chapter. The third part he intends he his intention is to write an order description of the movie.

0:15:7.937 --> 0:15:19.127 gianmaria.greco@unimc.it What he does, what Gregory Frazier, Frazier does. The very first thing is sits in a room with the blind and non blind people that have no knowledge of the movie.

0:15:19.847 --> 0:15:26.667 gianmaria.greco@unimc.it I'll never seen that movie, never watched it, and they listen to the movie for the very first time.

0:15:27.567 --> 0:15:35.17 gianmaria.greco@unimc.it And then they take notes on things that were not. They were not able to understand because they were listening and not watching it.

0:15:35.747 --> 0:15:38.177 gianmaria.greco@unimc.it And from all these notes and the discussion.

0:15:39.377 --> 0:15:58.207 gianmaria.greco@unimc.it So from there, gaps and interpretations or misinterpretation then that all that becomes became the starting point from Fraser to write the order description. So there is, I mean, at the very beginning or scholarly.

0:15:58.567 --> 0:16:6.477 gianmaria.greco@unimc.it A speaking about description accessibility was there was users were there and so on.

0:16:7.27 --> 0:16:13.117 gianmaria.greco@unimc.it Uh, so I'm not saying that's a it's something new, but.

0:16:13.807 --> 0:16:15.697 gianmaria.greco@unimc.it That the Frasers.

0:16:16.827 --> 0:16:21.507 gianmaria.greco@unimc.it My approach became for different reasons.

0:16:22.7 --> 0:16:32.217 gianmaria.greco@unimc.it Ohh, it's all really important to get into this reasons but became a not the dominant 1.

0:16:33.247 --> 0:16:47.987 gianmaria.greco@unimc.it OK. Anyway, so that's more or less. Where are we in a way in some of the things where are we talking about media accessibility? And it's not a precise map. Ohh.

0:16:49.377 --> 0:17:3.557 gianmaria.greco@unimc.it So I'm talking to you from the University of Surrey in Guildford, where I'm I'm spending a month as a visiting researcher and Guildford has a local hero which is Alan Turing.

0:17:4.557 --> 0:17:16.467 gianmaria.greco@unimc.it Umm, so that's why I'm international hero obviously, and this is the statue of Turing right here on campus. So ohh.

0:17:18.537 --> 0:17:23.347 gianmaria.greco@unimc.it In 19 during his one considered to be one of the.

0:17:23.927 --> 0:17:27.227 gianmaria.greco@unimc.it A pioneers and founders of.

0:17:28.47 --> 0:17:29.407 gianmaria.greco@unimc.it Artificial intelligence.

0:17:30.657 --> 0:17:58.537 gianmaria.greco@unimc.it Ohh no they article that I ask you to read the possibly. I will only mention at the very general thing about that article, but then if you wish the article is plenty, has plenty of of insights of aspects that could be linked and discussed within media accessibility so we can do that during the second part of the discussion now the.

0:17:59.547 --> 0:18:13.557 gianmaria.greco@unimc.it Or or the article or the title of the article is computing machinery and intelligence there was published in mind, which is one of the most prestigious journals in philosophy. In 1950 he starts.

0:18:14.197 --> 0:18:31.467 gianmaria.greco@unimc.it With the steering starts from the question, can machines think? And as you rent it refuses. He rejects the question, he refuses to answer the question for the city of reasons, but the core of it is that the.

0:18:31.547 --> 0:18:40.537 gianmaria.greco@unimc.it So ohh my terms machine and thinking are too vague of is that the devices?

0:18:41.217 --> 0:18:50.197 gianmaria.greco@unimc.it A situation a game which he called so the mutation game ohh he never called.

0:18:50.987 --> 0:18:53.67 gianmaria.greco@unimc.it Calls it a test.

0:18:54.397 --> 0:19:9.757 gianmaria.greco@unimc.it Not in that sense that we used during test today. It's a game for a for a Turing, but over the years that's how it has become famous. As the Turing test now.

0:19:11.577 --> 0:19:25.647 gianmaria.greco@unimc.it Again, I'm assuming that you read that article, but just in case, let me summarize The Imitation Game. Ohh, there are three players.

0:19:26.507 --> 0:19:55.937 gianmaria.greco@unimc.it One AB and C AM Br human beings. They are at the beginning of the game. They are in two separate rooms, so season is in a third room and they interact. See interacts with A and separately with B through let's say a computer or a paper or whatever. Let's say a computer, a computer interface, so a keyboard and a screen. And this is to avoid any bias.

0:19:56.507 --> 0:19:57.337 gianmaria.greco@unimc.it Ohh.

0:19:58.227 --> 0:20:8.877 gianmaria.greco@unimc.it He communicates, ask US questions way a replies as other questions to be be replies and.

0:20:11.217 --> 0:20:13.917 gianmaria.greco@unimc.it Shooting says OK if after a while.

0:20:14.977 --> 0:20:15.667 gianmaria.greco@unimc.it See.

0:20:16.477 --> 0:20:17.177 gianmaria.greco@unimc.it The one who's?

0:20:18.37 --> 0:20:26.577 gianmaria.greco@unimc.it Uh, asking the questions is not able to. Sorry I just forgot. Then I one point A or B are.

0:20:27.387 --> 0:20:28.627 gianmaria.greco@unimc.it Or you put.

0:20:30.27 --> 0:20:34.247 gianmaria.greco@unimc.it Hey, computer, artificial intelligence instead of either A&B.

0:20:35.447 --> 0:20:45.497 gianmaria.greco@unimc.it Now, if the map after a while the let's say the judge, the evaluator ohh does cannot identify.

0:20:46.187 --> 0:20:51.617 gianmaria.greco@unimc.it Which one is the AI and which one is the human being then?

0:20:52.707 --> 0:20:57.387 gianmaria.greco@unimc.it The AI the machine has passed us.

0:20:59.627 --> 0:21:4.587 gianmaria.greco@unimc.it That today they will say as passed the test, so one wins the game.

0:21:5.987 --> 0:21:7.637 gianmaria.greco@unimc.it Because at the.

0:21:8.537 --> 0:21:11.327 gianmaria.greco@unimc.it Ohm afford another? You want being?

0:21:12.37 --> 0:21:20.997 gianmaria.greco@unimc.it It cannot be a the the the another human being cannot distinguish which one is the AI and B now.

0:21:23.227 --> 0:21:32.657 gianmaria.greco@unimc.it That are again tons of things to say in terms of media accessibility and diversity and so on in this case. But first of all, it's.

0:21:35.397 --> 0:21:42.447 gianmaria.greco@unimc.it Don't focus on the linguistic aspects, OK? The the that is a specific case.

0:21:43.197 --> 0:21:50.267 gianmaria.greco@unimc.it Ohh but ohh. It's the idea that a behavior.

0:21:51.857 --> 0:22:4.797 gianmaria.greco@unimc.it That we assume to be inherently human. If a machine can do it, then what? During tells us basically is that either we.

0:22:5.497 --> 0:22:8.837 gianmaria.greco@unimc.it Decide we accept that that machine.

0:22:9.627 --> 0:22:10.907 gianmaria.greco@unimc.it Is displaying.

0:22:12.367 --> 0:22:24.587 gianmaria.greco@unimc.it Peeters that so far we thought they were inherently human or we have to change the definition of those features. Let me give you an example because in other.

0:22:24.887 --> 0:22:32.367 gianmaria.greco@unimc.it Ohh papers. During the uses the examples of of of.

0:22:33.807 --> 0:22:50.427 gianmaria.greco@unimc.it Playing chess, for example. Well, let's assume a for a long time. We assumed that playing chess was a a, an activity that required it telligence in order to be to play chess, you need to be intelligent.

0:22:51.137 --> 0:22:51.857 gianmaria.greco@unimc.it Ohh.

0:22:53.227 --> 0:22:53.917 gianmaria.greco@unimc.it Then.

0:22:55.37 --> 0:23:1.377 gianmaria.greco@unimc.it People created their lab and computer and AI and software. There was able to play.

0:23:2.677 --> 0:23:18.387 gianmaria.greco@unimc.it Ohh to play chess so we raise the bar. We changed OK chess in order to play chess. It's not. It's not just playing chess. So in order to play it as a professional.

0:23:19.487 --> 0:23:32.837 gianmaria.greco@unimc.it Again, they invented a software able to play at the level of a professional. No, we changed day, let's say changed again. The raised again the bar, the.

0:23:36.267 --> 0:23:46.377 gianmaria.greco@unimc.it The point is having a A for a, a laptop to be considered intelligent for a laptop, for a software to be considered intelligent is.

0:23:47.277 --> 0:23:52.447 gianmaria.greco@unimc.it Being able to win against a world champion.

0:23:53.107 --> 0:23:55.777 gianmaria.greco@unimc.it Done. Checked now.

0:23:56.807 --> 0:23:58.697 gianmaria.greco@unimc.it The whole point here is that.

0:23:59.607 --> 0:24:6.887 gianmaria.greco@unimc.it It's not that the machine per se is intelligent. Whatever intelligence means, it's that either.

0:24:7.687 --> 0:24:11.427 gianmaria.greco@unimc.it If the machine is able to display that behavior.

0:24:13.7 --> 0:24:14.607 gianmaria.greco@unimc.it A behavior that we assume.

0:24:15.247 --> 0:24:16.867 gianmaria.greco@unimc.it Require the intelligence.

0:24:17.587 --> 0:24:37.57 gianmaria.greco@unimc.it Very vaguely, this vague idea of intelligence, then either we accepted there is intelligent or intelligence. Intelligence is something else. The machine playing chess at the very proficient level doesn't require intelligence because that software is not intelligent. If we in they intuitive sense.

0:24:38.297 --> 0:24:52.787 gianmaria.greco@unimc.it Now all that said, just do it to give you a specific a bit an idea of what really the test is about. So it's not saying that the machine is intelligent, OK? It's really posing us a question.

0:24:58.387 --> 0:25:7.27 gianmaria.greco@unimc.it Now let's quickly have a look at this limitation game from categories that are more familiar to us.

0:25:8.127 --> 0:25:13.17 gianmaria.greco@unimc.it In the media accessibility, make your experts users now.

0:25:14.497 --> 0:25:22.17 gianmaria.greco@unimc.it BNC they two humans, no are are are there because?

0:25:23.157 --> 0:25:35.137 gianmaria.greco@unimc.it B. Is the human being, so it's supposed to be able to to have to possess some specific skills and expertise.

0:25:36.167 --> 0:25:50.297 gianmaria.greco@unimc.it As a maker, as a creator, OK of of in this case of semantic meaning and as a proficient user of language written language in the specific context.

0:25:51.567 --> 0:25:57.267 gianmaria.greco@unimc.it And A's trying to imitate B, but C is there.

0:25:58.687 --> 0:26:0.97 gianmaria.greco@unimc.it Because it's a user.

0:26:1.207 --> 0:26:6.927 gianmaria.greco@unimc.it So the judge actually is there. So the evaluator, those who decides.

0:26:7.707 --> 0:26:12.697 gianmaria.greco@unimc.it If a basically park wins the game or not, it's the user.

0:26:13.657 --> 0:26:14.887 gianmaria.greco@unimc.it So it's the user.

0:26:15.747 --> 0:26:19.37 gianmaria.greco@unimc.it Because the user has the one they is at the.

0:26:20.517 --> 0:26:31.817 gianmaria.greco@unimc.it Of the recipient and of a social interaction in this case. So the one that uses what the other is telling to that person.

0:26:33.657 --> 0:26:41.667 gianmaria.greco@unimc.it So that's it. Very interesting because in The Imitation Game, that decision whether the A passes or not.

0:26:42.507 --> 0:26:43.617 gianmaria.greco@unimc.it Winds or not?

0:26:44.417 --> 0:27:11.7 gianmaria.greco@unimc.it Uh is based on the user. It's the it's the judge is there because it's a user of language. Since it uses then it cannot see if the two the replies from AMB are coherent or not are valid or not and so. So now let's forget about this. Then we will try to come back this.

0:27:13.337 --> 0:27:21.667 gianmaria.greco@unimc.it In a few minutes afterwards, the end actually, and let's move on. The human technology continue so.

0:27:24.67 --> 0:27:51.357 gianmaria.greco@unimc.it Love around the 1975 five I philosopher of technology called the Don either published the book and the before that also some articles where we started to discuss about the human human technology, human machine relations, talking about the continuum now is continuum is a completely different answer from the one that I'm gonna talk about is very relevant for us. It's not the topic of the day.

0:27:52.737 --> 0:27:53.907 gianmaria.greco@unimc.it It's not the way.

0:27:55.807 --> 0:27:56.857 gianmaria.greco@unimc.it People in.

0:27:56.947 --> 0:28:7.467 gianmaria.greco@unimc.it Are are, are usually use that the labeled expression that human technology continue or human machine continue either.

0:28:7.547 --> 0:28:26.517 gianmaria.greco@unimc.it So uh talked about talks about a continuum between technology as transparent as a kind of almost invisible as part inherent part of our lives. And on the other hand, the technology has kind of otherness and other compared to us.

0:28:27.497 --> 0:28:34.947 gianmaria.greco@unimc.it But over time, the either probably this is just my my.

0:28:38.187 --> 0:28:55.117 gianmaria.greco@unimc.it Assumption based on what I read over the years, so either probably was the one who or, or mainstreamed, or after either, or the expression of human technology continuum. You know, machine continuum become became mainstreamed but.

0:28:56.787 --> 0:29:4.667 gianmaria.greco@unimc.it With the different with a different meaning and the meaning they still used today, but was changed pretty pretty soon.

0:29:7.567 --> 0:29:19.437 gianmaria.greco@unimc.it Is a quote from Folds 1999. This is a paper about avionics. Something completely different is not talking about really specific things in avionics by one point.

0:29:19.937 --> 0:29:34.277 gianmaria.greco@unimc.it Ohh, he says. Yeah. And there is a continuum. The range is from entirely performed by the human to entirely performed by the machine and in the middle of this continuum, functions are performed jointly by human and machine.

0:29:35.77 --> 0:29:45.287 gianmaria.greco@unimc.it Now what these human technology, human machine continuum look like in a way. So on the one hand we have.

0:29:47.707 --> 0:29:51.687 gianmaria.greco@unimc.it Now actions, activities that there are are performed.

0:29:52.547 --> 0:29:54.777 gianmaria.greco@unimc.it Ohh only by a human being.

0:29:56.537 --> 0:29:57.957 gianmaria.greco@unimc.it Then we have.

0:29:59.17 --> 0:30:12.907 gianmaria.greco@unimc.it Those that are human centered, so there's still the human that's doing the core of it, but they are technology assistant and then those that are technology centers or technologies doing.

0:30:13.647 --> 0:30:40.147 gianmaria.greco@unimc.it The core of it and the core of the function and, but there is the human interloop somewhere, and finally the on the the other end, those functions, those activities, those services and so on. There are technology only based, they are fully automated. There is no intervention of a human in the loop.

0:30:41.437 --> 0:30:50.787 gianmaria.greco@unimc.it There's no human in the loop, but if not just the one who wrote the software at the very beginning. But that's it. The action part is not performed by.

0:30:51.147 --> 0:30:54.247 gianmaria.greco@unimc.it Ohh has not human in the loop as they say.

0:30:55.797 --> 0:31:25.267 gianmaria.greco@unimc.it Now, this idea of human technology continuum, as I said, it's not new. It's not even a Edens, but is a we can find it, even including in some in some of his writing, even though not express with the label human technology continuum. But we can find it in a Barry, a very nice, brilliant book.

0:31:25.427 --> 0:31:36.977 gianmaria.greco@unimc.it Published by note, but Vienna, who is the pioneer of the founder of Sabermetrics, published the very same year in 1955. So same year touring published this paper.

0:31:38.87 --> 0:31:48.597 gianmaria.greco@unimc.it And ohh in in that book is speaks. It doesn't use the the human technology continuum.

0:31:48.697 --> 0:31:58.107 gianmaria.greco@unimc.it A that expression, but it talks about this kind of continue and at one point is as yeah. And along this continuum.

0:32:0.227 --> 0:32:1.207 gianmaria.greco@unimc.it Machines.

0:32:2.147 --> 0:32:5.197 gianmaria.greco@unimc.it And when he says machines is not talking just about.

0:32:6.617 --> 0:32:28.177 gianmaria.greco@unimc.it Ohm something. Some hardware is putting intelligent machines OK? AI in the mix as well. They can be also used for example, to make up for the losses of the maimed and of the sensory deficient. Now obviously it's 1950. So the.

0:32:28.837 --> 0:32:31.147 gianmaria.greco@unimc.it Had Menology is medical model.

0:32:31.987 --> 0:32:37.897 gianmaria.greco@unimc.it But what matters here is that there was already an idea.

0:32:38.947 --> 0:32:41.567 gianmaria.greco@unimc.it Uh, and it's not very new that.

0:32:43.387 --> 0:32:54.697 gianmaria.greco@unimc.it Technology and the different ratio between the human and technology. We can create technologies that along that ratio can provide can be.

0:32:55.217 --> 0:32:58.517 gianmaria.greco@unimc.it A what we call today access services.

0:33:3.297 --> 0:33:13.547 gianmaria.greco@unimc.it Again, please forget the medical model terminology. Just abstract and contextualizes in. Contextualize that in the 1950.

0:33:14.337 --> 0:33:16.487 gianmaria.greco@unimc.it So what this could look like?

0:33:17.537 --> 0:33:23.67 gianmaria.greco@unimc.it Oh, oh, in media accessibility, for example.

0:33:24.187 --> 0:33:24.577 gianmaria.greco@unimc.it Well.

0:33:25.507 --> 0:33:34.757 gianmaria.greco@unimc.it P to P, persons to person, not all your scription. I owe these P to P to annamjankowska. I struggled a bit to find that this acronym.

0:33:36.997 --> 0:33:38.577 gianmaria.greco@unimc.it Me all the describing.

0:33:39.367 --> 0:33:45.677 gianmaria.greco@unimc.it Something to someone else just in person, someone sitting next to me is purely human.

0:33:46.587 --> 0:34:2.967 gianmaria.greco@unimc.it Ohh respeaking, at least at the moment is the color of the activities done by human based technology assisted at the moment and think about some forms of post edited services especially.

0:34:3.777 --> 0:34:11.747 gianmaria.greco@unimc.it Those who are increasingly becoming very good, so the core of the activities performed by technology, but then.

0:34:12.167 --> 0:34:31.57 gianmaria.greco@unimc.it A. There is some activity that post editing done by the human and think about audio subtitling or an not real yet audio auto audio description. Those will be technology only fully automated services.

0:34:32.887 --> 0:34:33.537 gianmaria.greco@unimc.it But.

0:34:34.227 --> 0:34:53.67 gianmaria.greco@unimc.it It's not only about within this range, there is also another aspect that we should take into account. I'm not so sure why there is that thing and is the intensity, so it's not just how much.

0:34:54.427 --> 0:34:59.657 gianmaria.greco@unimc.it He's done by the how much of the activity?

0:35:0.367 --> 0:35:2.647 gianmaria.greco@unimc.it In the sense of which one is the core.

0:35:3.147 --> 0:35:9.437 gianmaria.greco@unimc.it Uh. The core component is either technology or human, but how much is intense?

0:35:10.507 --> 0:35:13.457 gianmaria.greco@unimc.it Ohh for example all automated.

0:35:13.527 --> 0:35:16.667 gianmaria.greco@unimc.it OHS, OHS subtitles. So ideally.

0:35:17.377 --> 0:35:37.317 gianmaria.greco@unimc.it Are completely technology intensive, but zero human intensive in the sense of the creation. OK, that's what I'm saying. On the other end, you may know the description, the P2P or the description mentioned before has nothing about technology. It's entirely human and within these.

0:35:37.387 --> 0:35:41.797 gianmaria.greco@unimc.it Oh oh. Or area there is a continuous shift.

0:35:43.627 --> 0:36:0.227 gianmaria.greco@unimc.it I'll come back to this later, so talk to you later now. So there is a constant interaction and a series of shifts are both along the human technology continuum and the human technology intensity in terms of human technology intensity. And this means that.

0:36:1.177 --> 0:36:2.427 gianmaria.greco@unimc.it The power relation.

0:36:3.127 --> 0:36:5.197 gianmaria.greco@unimc.it Between now, the human.

0:36:5.837 --> 0:36:6.957 gianmaria.greco@unimc.it And the machine.

0:36:8.197 --> 0:36:9.887 gianmaria.greco@unimc.it Are constantly shifting.

0:36:11.487 --> 0:36:14.657 gianmaria.greco@unimc.it And then this will pose us some questions. OK.

0:36:16.707 --> 0:36:20.87 gianmaria.greco@unimc.it Now moving towards the.

0:36:21.447 --> 0:36:24.347 gianmaria.greco@unimc.it Opening the second part of let's say Ohh.

0:36:27.177 --> 0:36:36.657 gianmaria.greco@unimc.it A few decades ago, talking about mass media communication and especially mass meet cultural produced by mass media communication Umberto Eco.

0:36:37.217 --> 0:36:37.857 gianmaria.greco@unimc.it Ohh.

0:36:40.627 --> 0:37:10.877 gianmaria.greco@unimc.it Stad describe the possibility to of of divide in the case where the intellectuals or into the apocalyptic apocalyptic ones. Those were completely against the mass media communication. That's it, full stop. No, no way, and not the integrated ones. Those who are so much part of the mechanism that they didn't even realize they were part of the mechanism were just doing it.

0:37:11.537 --> 0:37:19.507 gianmaria.greco@unimc.it Now, what about these cantaloupes about AI and then later about AI and media accessibility?

0:37:21.507 --> 0:37:23.127 gianmaria.greco@unimc.it So if we look at the.

0:37:23.877 --> 0:37:24.817 gianmaria.greco@unimc.it Attitudes.

0:37:25.997 --> 0:37:33.847 gianmaria.greco@unimc.it The words AI in general, I'm not talking about media accessibility, diversity in general, about artificial intelligence in society.

0:37:34.607 --> 0:37:42.137 gianmaria.greco@unimc.it We can see that we have the apocalyptic those who say no way, no way. It's dangerous, so we avoided.

0:37:44.27 --> 0:37:45.997 gianmaria.greco@unimc.it And those were just do it.

0:37:46.827 --> 0:38:3.497 gianmaria.greco@unimc.it They just write their own love their software, they do it, they don't ask questions. They don't think about social effects and and so on. Ohh and those who are enthusiasts actually. So we need to amplify a bit to expand that.

0:38:4.137 --> 0:38:9.407 gianmaria.greco@unimc.it Ohh or simplify the model or?

0:38:10.117 --> 0:38:17.887 gianmaria.greco@unimc.it Device by echo because there are also the enthusiasts. No, no, it will solve all the problems of humanity.

0:38:18.857 --> 0:38:32.87 gianmaria.greco@unimc.it Ohh, kind of. If you like the distopian future, the apocalyptic on the one hand, the utopian future on. On the other hand, in the middle, those who just do it without even asking questions.

0:38:33.127 --> 0:38:49.707 gianmaria.greco@unimc.it But if we look closely at at not so closely actually doesn't really take a lot of effort to look at the debate. There are not just these three roads. There's a a fourth row which challenge calling the watchful I don't.

0:38:50.407 --> 0:38:53.477 gianmaria.greco@unimc.it Of the vigilant, but I think watchful it's less. It's nice.

0:38:53.997 --> 0:38:58.827 gianmaria.greco@unimc.it Uh, I'm open to suggestions. Those who says no.

0:38:59.807 --> 0:39:0.457 gianmaria.greco@unimc.it What?

0:39:1.227 --> 0:39:13.967 gianmaria.greco@unimc.it These are these are interesting technologies. AI can really change the the way can really change society.

0:39:15.247 --> 0:39:18.717 gianmaria.greco@unimc.it But we need to be aware and watch.

0:39:19.757 --> 0:39:21.767 gianmaria.greco@unimc.it Ohh, we need to be vigilant.

0:39:22.567 --> 0:39:30.337 gianmaria.greco@unimc.it Which means that they raise since the beginning. They are raising now the discussion.

0:39:31.207 --> 0:39:33.997 gianmaria.greco@unimc.it To a social level, so before.

0:39:35.7 --> 0:39:39.37 gianmaria.greco@unimc.it The A of a massive.

0:39:39.747 --> 0:39:47.117 gianmaria.greco@unimc.it Development of mainstream of the of this specific technology they are calling for a.

0:39:47.767 --> 0:39:53.317 gianmaria.greco@unimc.it Discussion about the social ethical consequences, the watchful.

0:39:54.927 --> 0:40:3.467 gianmaria.greco@unimc.it Now I'm asking you, I will where is media accessibility? Where are we as a media accessibility?

0:40:4.507 --> 0:40:5.237 gianmaria.greco@unimc.it I'm almost done.

0:40:7.447 --> 0:40:10.277 gianmaria.greco@unimc.it The media accessibility is not in this debate.

0:40:11.307 --> 0:40:15.777 gianmaria.greco@unimc.it If you go outside and read the.

0:40:17.637 --> 0:40:33.577 gianmaria.greco@unimc.it Technical papers and the debate uh media accessibility. I'm talking in this case about us. It's not that it's not in the in that debate. Our voices are not there. We are.

0:40:34.247 --> 0:40:38.457 gianmaria.greco@unimc.it At the moment there are some people there are doing.

0:40:39.347 --> 0:40:40.817 gianmaria.greco@unimc.it So there are more integrated.

0:40:41.577 --> 0:40:42.257 gianmaria.greco@unimc.it Ohh.

0:40:43.37 --> 0:41:13.887 gianmaria.greco@unimc.it Some are apocalyptic, where some most are apocalypse apocalyptic was I participated last November in the conference, the European, the Translating Europe Forum, organized by the digit translation, and for two days or three days. I don't remember. The was kind of ghost going around, which was machine translation which should be stopped. The people were really crazy.

0:41:13.967 --> 0:41:15.977 gianmaria.greco@unimc.it There was this feeling they should be stopped.

0:41:16.617 --> 0:41:17.227 gianmaria.greco@unimc.it Ohh.

0:41:18.17 --> 0:41:20.727 gianmaria.greco@unimc.it But there are not that many enthusiasts.

0:41:21.677 --> 0:41:29.727 gianmaria.greco@unimc.it But there are almost none watchful there is not a debate. OK, before start we start doing this.

0:41:30.487 --> 0:41:34.487 gianmaria.greco@unimc.it We need to raise the discussion to a different level.

0:41:35.587 --> 0:41:36.107 gianmaria.greco@unimc.it Now.

0:41:38.877 --> 0:41:53.897 gianmaria.greco@unimc.it In moving towards the conclusion now the great, let's go back to the recent past just to say that all those that debate towards the importance of access for everyone and the fact that.

0:41:54.387 --> 0:42:10.387 gianmaria.greco@unimc.it A that requires an attention to our own the the diversity that is part of each human being, the importance of users and our users should be involved since they they one and so on and so on.

0:42:11.467 --> 0:42:12.177 gianmaria.greco@unimc.it These.

0:42:12.867 --> 0:42:19.187 gianmaria.greco@unimc.it Despite the historically they we had cases like Frasers.

0:42:20.737 --> 0:42:22.727 gianmaria.greco@unimc.it But they are very recent recent.

0:42:23.757 --> 0:42:26.877 gianmaria.greco@unimc.it Which means that they are weak.

0:42:28.17 --> 0:42:29.587 gianmaria.greco@unimc.it They are still weak.

0:42:32.167 --> 0:42:32.707 gianmaria.greco@unimc.it Uh.

0:42:33.987 --> 0:42:34.517 gianmaria.greco@unimc.it And.

0:42:35.307 --> 0:42:38.837 gianmaria.greco@unimc.it Which means that AI could.

0:42:40.67 --> 0:42:40.737 gianmaria.greco@unimc.it Improve.

0:42:42.627 --> 0:42:45.957 gianmaria.greco@unimc.it To the bridge. What for a time.

0:42:46.697 --> 0:42:59.307 gianmaria.greco@unimc.it Within media accessibility, that's what I'm talking about. Has been a epistemic divide, which is still there, right. But now we are aware of this. So at discrimination of people as nowheres.

0:43:0.337 --> 0:43:32.147 gianmaria.greco@unimc.it And also poietic divide discrimination of people as makers. Now AI can bridge this divide or it cannot extend and strengthen this divide, make it wider. Ohh, let me quote the again that book. The book, written by Norbert Wiener. At one point he says that yeah these machines maybe used to make up for the losses of the name and the sensory division. That was the first part.

0:43:32.447 --> 0:43:46.877 gianmaria.greco@unimc.it But so they can as well as to give new and potentially dangerous powers to the already powerful, powerful, and increase the bias the discrimination within society now.

0:43:47.517 --> 0:43:50.247 gianmaria.greco@unimc.it Ohh as a way of concluding.

0:43:53.937 --> 0:43:55.597 gianmaria.greco@unimc.it Let me give you this example.

0:43:56.447 --> 0:43:59.387 gianmaria.greco@unimc.it The OHH at 1.1 more.

0:44:0.207 --> 0:44:2.117 gianmaria.greco@unimc.it Sorry, I'm really trying to.

0:44:4.407 --> 0:44:15.597 gianmaria.greco@unimc.it Summarize it. Ohh you know that one point people were working on the creation of a software able to be used through a mobile phone and.

0:44:15.667 --> 0:44:21.827 gianmaria.greco@unimc.it So in order for it to be used by blind people so that they could then.

0:44:22.107 --> 0:44:48.567 gianmaria.greco@unimc.it My frame with their phone and object and that object will be recognized and read out loud or recognized by the software. So at the very beginning these the database of the images to train the software, we're taken by people from the company, sighted people, OK, not non blind people. And there's all the failed. And then they blindfolded themselves.

0:44:49.767 --> 0:45:5.747 gianmaria.greco@unimc.it And again being blindfold, blindfolding myself to simulate a blind person of the blind person takes a picture of and then feeding that picture into the software is not the same thing. So finally they created a.

0:45:6.67 --> 0:45:25.567 gianmaria.greco@unimc.it Ohh database. Let's say a corpus of pictures taken by blind people that were fed into the software and the software works much better. What is the problem that they were doing it from a non user centering, no proactive and so on perspective. There is a risk now.

0:45:26.227 --> 0:45:29.7 gianmaria.greco@unimc.it So first of all, I think that we need to be careful.

0:45:31.47 --> 0:45:43.357 gianmaria.greco@unimc.it Yes, we need to or AI here. We cannot stop. It doesn't really make sense to stop it, but we need to be careful. And I think that there is a risk that we start.

0:45:43.937 --> 0:45:45.547 gianmaria.greco@unimc.it Ohh using it.

0:45:47.517 --> 0:45:53.667 gianmaria.greco@unimc.it Yeah, we started using it, doing research on it. Remember the story.

0:45:54.477 --> 0:46:7.987 gianmaria.greco@unimc.it That I told you at the beginning. Now, first of all, we start. There is a risk. We start doing research and using it for getting this kind of sheets. So there is a risk of going back.

0:46:9.197 --> 0:46:31.427 gianmaria.greco@unimc.it Facades, or at least years, and I'll tell you another story. Another time. But then thinking about the story I told you at the beginning, it's that story. What matters to me is that Bob, you remember Bob after reading just three pages, started doing a presentation about quantum mechanics. But OP is an architect.

0:46:32.7 --> 0:47:2.67 gianmaria.greco@unimc.it Ohh is that there is a risk to start talking about AI without having proper solid knowledge or where I just because it's cool. It's something that let's be honest. We experience with experimental research in a in ABT for an media accessibility for some time at one point we start doing it without really knowing how I thesis testing had to be done and there are papers.

0:47:2.137 --> 0:47:19.97 gianmaria.greco@unimc.it Let's say things that that will never be published in a scientific journal because there were no scientific at all. They were saying things that started from a, for example, new hypothesis testing are not conceivable, are not really not possible. So we need to.

0:47:19.597 --> 0:47:33.727 gianmaria.greco@unimc.it Ohh not only to be careful not to lose what we gained so far, at least in terms of awareness and debate, but also to be solid and finally to be watchful. Ohh the.

0:47:34.617 --> 0:47:56.207 gianmaria.greco@unimc.it AI in media accessibility is really at the beginning. The discussion is really the beginning. There has been some project here in Guildford, for example, but there is not that much and something about automatic subtitles, but the whole core of research on AI media accessibility is.

0:47:56.287 --> 0:48:9.297 gianmaria.greco@unimc.it Ohh has to come yet. OK, so I think this is really the right point. We should start to become watchful and to have.

0:48:11.307 --> 0:48:17.97 gianmaria.greco@unimc.it To have this kind of view on what are we aiming for.

0:48:17.837 --> 0:48:47.447 gianmaria.greco@unimc.it Ohh this AIM a hopefully we are aiming for good. Something good we are and we need to be aware of this. That's the whole point and I think that you ask or doctoral students and a doctoral researchers and postdoctoral researchers. You are really the ones we should place our hop hope ohh. Our hope in having a.

0:48:48.47 --> 0:48:57.17 gianmaria.greco@unimc.it In developing a area research AI media accessibility, there is solid, careful and watchful. Thank you very much.

0:49:2.857 --> 0:49:4.307 Alicja Zajdel I think kids and Maria.

0:49:5.247 --> 0:49:6.147 Alicja Zajdel I think this is.

0:49:7.267 --> 0:49:7.577 Alicja Zajdel Yeah.

0:49:4.827 --> 0:49:16.47 gianmaria.greco@unimc.it It's so awkward. OK, I hope you were able to listen to me because I complete silence. It's really, really, really awkward.

0:49:18.407 --> 0:49:18.847 gianmaria.greco@unimc.it Sorry.

0:49:17.27 --> 0:49:35.417 Alicja Zajdel I know not having any feedback, but no, thank you. We we heard you perfectly. I'm personally not sure I understood everything. It will it. It's very, very, very conceptual level. But I was very interesting. I think this was a great place to start a series of guest lectures.

0:49:36.237 --> 0:49:51.427 Alicja Zajdel And and like you said yourself, it's a debate we don't really hear so much in in media accessibility. So everyone please, if you have a question, you can put it in the chat or you can just unmute yourself. You can put your camera on and yeah, please go ahead.

0:49:50.867 --> 0:49:52.797 gianmaria.greco@unimc.it Let me see that shot.

0:49:56.637 --> 0:49:57.327 gianmaria.greco@unimc.it Better jump.

0:50:4.107 --> 0:50:11.917 gianmaria.greco@unimc.it Don't be shy. Even if you don't. Didn't like something. Please. That's would be even good. Actually, even better.

0:50:12.877 --> 0:50:13.387 Kate Anne Dangerfield (Research Student) Hello.

0:50:12.857 --> 0:50:17.47 Alicja Zajdel And then you put your hand up. You can you unmute yourself. What do I have to do it?

0:50:19.747 --> 0:50:20.187 gianmaria.greco@unimc.it Alright.

0:50:18.57 --> 0:50:22.847 Sabien Hanoulle (Guest) I can do it. Hi. Yeah, I saw eight. Raised her real hand.

0:50:24.227 --> 0:50:28.357 Sabien Hanoulle (Guest) So well, to give you the start, Jan Maria.

0:50:29.307 --> 0:50:32.307 Sabien Hanoulle (Guest) I'm not sure whether you're talking to the right people.

0:50:34.237 --> 0:50:38.67 Sabien Hanoulle (Guest) I fully agree with what you're saying, but.

0:50:40.427 --> 0:50:44.537 Sabien Hanoulle (Guest) And we had the same feeling at the translating Europe Forum.

0:50:45.257 --> 0:50:47.787 Sabien Hanoulle (Guest) Uh, I think we should.

0:50:50.367 --> 0:50:52.137 Sabien Hanoulle (Guest) Talk with the industry.

0:50:53.237 --> 0:50:58.397 Sabien Hanoulle (Guest) What the what you say? Make them more watchful because they are.

0:51:0.267 --> 0:51:2.957 Sabien Hanoulle (Guest) That's my feeling As for what I.

0:51:4.177 --> 0:51:4.917 Sabien Hanoulle (Guest) Know about it.

0:51:5.817 --> 0:51:9.287 Sabien Hanoulle (Guest) I feel that the industry is pushing very much and not.

0:51:10.37 --> 0:51:10.837 Sabien Hanoulle (Guest) Umm.

0:51:11.667 --> 0:51:12.517 Sabien Hanoulle (Guest) Paying.

0:51:14.457 --> 0:51:23.287 Sabien Hanoulle (Guest) Enough attention to the quality of where we apply artificial artificial intelligence in media accessibility.

0:51:24.497 --> 0:51:27.47 Sabien Hanoulle (Guest) And and we are the ones saying.

0:51:28.227 --> 0:51:36.677 Sabien Hanoulle (Guest) Uh, wait a little. Let's we don't want a 90% quality, we want a 99% quality.

0:51:37.497 --> 0:51:38.287 Sabien Hanoulle (Guest) Umm.

0:51:39.77 --> 0:51:41.267 Sabien Hanoulle (Guest) Where we use artificial intelligence.

0:51:42.267 --> 0:51:47.557 Sabien Hanoulle (Guest) But on the other hand, it is very hard to talk with the industry because they only.

0:51:48.597 --> 0:51:49.717 Sabien Hanoulle (Guest) Get their money.

0:51:50.377 --> 0:51:57.537 Sabien Hanoulle (Guest) So I it's I think it's very difficult. I just wanted your thoughts about this.

0:51:56.77 --> 0:52:6.887 gianmaria.greco@unimc.it Yeah. Ohh, let me take this so I can write down the the things for digital and let's go on analog.

0:52:7.427 --> 0:52:37.197 gianmaria.greco@unimc.it Uh, that ohh? First of all, yeah, I I do get your point and. But I personally think I hope the OR at least this is the audience that wish to speak to, to say what I said, meaning that these my point is not going. I'm not talking about going to the industry and saying don't do this. Don't do that. There is a part of the discussion. But what I'm worried about at the moment, OK.

0:52:38.87 --> 0:53:4.997 gianmaria.greco@unimc.it Is that we start all that. There is this thing called AI. Let's start doing an experiment or a project about AI in an audio description. I don't know I and and we start developing a new forms and we forget that we need to involve users. We need that since we have a new tool, we start doing projects and and doing research for getting where everything we discussed how we matured.

0:53:6.927 --> 0:53:37.877 gianmaria.greco@unimc.it Uh, as a field? I'm talking right now as a scholars, OK, as scholars as researchers. And that's why I want to talk to doctoral researchers and postdoctoral researchers because they are the ones that, let's face it, whatever they are gonna do at one point, they will in the next years, they probably will face a I somehow. So I want them to. That's my opinion. Of course, I wish. I don't want you to do whatever you want. It's a kind of personal wish that not to forget.

0:53:38.537 --> 0:53:40.827 gianmaria.greco@unimc.it What we had to fight for in a way, OK?

0:53:41.947 --> 0:53:58.567 gianmaria.greco@unimc.it And the start doing everything from scratch. The example of the software for the recognition that is really a case. So let's create a software that can help in that case. That was the intention. OK, that can help blind people to.

0:53:58.907 --> 0:54:5.817 gianmaria.greco@unimc.it Ohh frame and object and the software will say OK, this is a cup OK and.

0:54:6.37 --> 0:54:7.507 gianmaria.greco@unimc.it A.

0:54:9.467 --> 0:54:13.667 gianmaria.greco@unimc.it Ohh the this is a cup and sorry I just received the.

0:54:14.927 --> 0:54:22.587 gianmaria.greco@unimc.it Ohh Ania, stop texting me like nice things to try things. Nice things. Sorry. I'm saying it public.

0:54:23.7 --> 0:54:52.877 gianmaria.greco@unimc.it Uh, the was saying the so they they this but that database, the first one of the images that they fed into the machine that the algorithm were taken by non blind people the way a non blind person uses a phone to frame an object having an it's completely different of what we do then they blindfold themselves again. It's I'm not blind so even blindfolded.

0:54:52.947 --> 0:55:22.967 gianmaria.greco@unimc.it I'm going to use differently and the machine was failing at. Then put it in the the the the recognition software was trained wrongly. They finally understood. OK, let's ask people blind people to take photos of. Let's feed that into the software and now the software is able to successfully to do it. It took them ten years to get this in a point. OK, so that's my point. Sabien. Let's not start to use something we did this kind of.

0:55:23.47 --> 0:55:27.47 gianmaria.greco@unimc.it Error if you like. It's a naive dievel with experimental research.

0:55:27.987 --> 0:55:41.807 gianmaria.greco@unimc.it There are papers. All the papers. OK, all papers. But publishing leading to published in leading journals by leading scholars. They say things. So there are completely unti scientific from the.

0:55:46.647 --> 0:55:47.977 Sabien Hanoulle (Guest) I get your points, yeah.

0:55:42.597 --> 0:55:57.177 gianmaria.greco@unimc.it I can give you an example because it would be too easy to Google it, but that's we'll never be, so let's not become biased in a way you know, not so being biased back, or at least be aware that there is my point.

0:55:56.667 --> 0:55:57.787 Sabien Hanoulle (Guest) OK. Thank you.

0:56:2.567 --> 0:56:5.807 gianmaria.greco@unimc.it Gate, you raise your hand. Hey.

0:56:6.177 --> 0:56:8.237 Kate Anne Dangerfield (Research Student) So hi, how are you?

0:56:10.267 --> 0:56:38.877 Kate Anne Dangerfield (Research Student) Thank you very much for the talk. It was really interesting I think I, yeah, I got quite a few questions, but I'll just go with one. But the I think what I'm thinking about is and maybe this is also relevant when it's about sort of academics and researchers, but it's like, you know, the impact that AI has on theory and you know these concepts that have been like from.

0:56:39.437 --> 0:57:6.867 Kate Anne Dangerfield (Research Student) And you know, all all I've sort of older concept from philosophy for instance that you know, they consider the human and and actually we're now in such a different position considering non humans and you know different relationships. And I just wonder really what well I think it's too big a question in lots of ways but it'd be interesting to hear what you think that you know what impact AI has on theory.

0:57:12.757 --> 0:57:14.707 Kate Anne Dangerfield (Research Student) That's different. That's just, yeah.

0:57:8.467 --> 0:57:15.527 gianmaria.greco@unimc.it OK, this is the. This is the only question is that they have many, so, OK, OK, OK, let's use this one.

0:57:15.907 --> 0:57:16.817 gianmaria.greco@unimc.it Ohm.

0:57:18.227 --> 0:57:21.477 gianmaria.greco@unimc.it What AI is having an effect on theory but.

0:57:21.557 --> 0:57:30.937 gianmaria.greco@unimc.it But that's why I introduced but that I really didn't have time to go into The Imitation Game and why it's important that we look into it.

0:57:32.397 --> 0:57:48.297 gianmaria.greco@unimc.it But Ohh, first of all, the mutation game it's a not a test for intelligent. That's why we sold. So the Turing test as a test for intelligence is not what is challenging us. It's saying that.

0:57:49.377 --> 0:58:1.987 gianmaria.greco@unimc.it We have to be. It's really interesting from a linguistic point, linguistic point of view for us and which means also, you know, the way we use terminology and so on and so on, because it's it.

0:58:3.627 --> 0:58:20.737 gianmaria.greco@unimc.it Intelligent. We believe that being able to drink from a cup A is something that requires necessarily, OK, necessarily, if you are intelligent, you. If you drink from a cup then it means that you're intelligent. That's then if you are able to.

0:58:21.257 --> 0:58:37.97 gianmaria.greco@unimc.it To to build a machine they is able to drink from a cup. Then the machine has to be intelligence or intelligence. Is something else. Which means if we believe that it really asking us questions about what it means to be human, because being able to play chess.

0:58:38.397 --> 0:58:45.427 gianmaria.greco@unimc.it For a long time, we thought it was something that you, it was necessarily being human. It was a necessary requirement.

0:58:46.187 --> 0:59:4.77 gianmaria.greco@unimc.it Necessarily means that as long as you are able to play Childs chess OK that you're human is able to play chess, then we build something. So playing chess is not can be played by humans, but it's not something that requires necessarily to be human.

0:59:4.857 --> 0:59:18.587 gianmaria.greco@unimc.it To to be performed. So there are it. It's helping us in a way to take out all the things that we thought we assume there were inherently human and only human.

0:59:19.707 --> 0:59:22.457 gianmaria.greco@unimc.it And that OK, look, all these things.

0:59:23.697 --> 0:59:28.307 gianmaria.greco@unimc.it Being able to write an essay on whatever it's not that probably.

0:59:29.377 --> 0:59:37.857 gianmaria.greco@unimc.it That human obviously with the with the caveat that he has all the kind of knowledge that knowledge base the OK.

0:59:38.607 --> 0:59:55.117 gianmaria.greco@unimc.it So what is the leaving? It's a asking us to or or to repurpose in a way the question what what is the defines us as humans and again put it within the disability and diversity. There is very interesting.

0:59:56.7 --> 1:0:15.487 gianmaria.greco@unimc.it Because for a long time we define persons with disabilities are not human because they were not able to do things. Some things now look, I'm not able to. I'm not able. I can do it but also machine can do it. So there is questioning my being human. Maybe I'm not the human. That's what distinguished me from someone else.

1:0:16.87 --> 1:0:26.237 gianmaria.greco@unimc.it Uh, so this is really changing. Now all this discussion needs to happen now at the sooner the better, because then the research.

1:0:27.587 --> 1:0:31.397 gianmaria.greco@unimc.it On the what I said, the researcher is automatic.

1:0:31.467 --> 1:0:48.337 gianmaria.greco@unimc.it Ohh subtitles. OK so there research or whatever whatever then is is happening within a ethical, social and theoretical debate. It's not just let's do something. Then we'll figure out this is good or not for society.

1:0:49.77 --> 1:0:58.597 gianmaria.greco@unimc.it Ohh so that's why I I take I change the agenda because agenda usually maybe interpreted as having a research agenda. I'm not saying that.

1:0:59.377 --> 1:1:16.697 gianmaria.greco@unimc.it You do the research on whatever you want, but let's I'm not want to do research on these and not on that. It's that the mind the the awareness of what, they're that. OK, let's do research on this, that direction. Take me on something. Please send me a while.

1:1:17.477 --> 1:1:21.877 gianmaria.greco@unimc.it But there is a a I don't know if it made sense, Kate.

1:1:22.727 --> 1:1:24.617 Kate Anne Dangerfield (Research Student) It does, yeah. Thank you very much.

1:1:27.287 --> 1:1:39.77 Alicja Zajdel As I lay, there is also a question in the chat. So we talked about all the possible risks associated with AI, but there is a question whether there are also some possible advances.

1:1:39.157 --> 1:1:50.297 Alicja Zajdel And the sorry possible benefits for media accessibility and more specifically considering the principle of personalization from your politic design.

1:1:51.187 --> 1:2:19.327 gianmaria.greco@unimc.it Yeah, that that is a one of the promises right of the AI. By again, we need to be aware of that and know the idea of the personalization such that each one of us is different. So it doesn't exist to one access. It doesn't exist. The one solution for all doesn't exist. The one ring to rule them. All, right. It's one of us has it's own ring to dominate its own society its own way.

1:2:19.567 --> 1:2:27.587 gianmaria.greco@unimc.it To to live its own life and needs access. So personalization needs something very important. Now AI can something that.

1:2:28.477 --> 1:2:37.857 gianmaria.greco@unimc.it Really can do that in theory. OK. And that's a debate in AI and disability, there is already happening, but.

1:2:38.867 --> 1:2:52.247 gianmaria.greco@unimc.it We need to to look at also the my. We need to have a conversation on on that as well. For example in personalization, personalization means also profiling people.

1:2:54.407 --> 1:3:2.197 gianmaria.greco@unimc.it Providing a personalized service through, I means that AI is somehow the software is profiling you to the dot.

1:3:3.37 --> 1:3:22.597 gianmaria.greco@unimc.it And what that information is going, how is going to be used in privacy terms in terms of privacy is something that needs to to be again discussed because it's will be brilliant, a word or Kim where?

1:3:23.657 --> 1:3:25.517 gianmaria.greco@unimc.it Really fantasy thing?

1:3:26.117 --> 1:3:55.637 gianmaria.greco@unimc.it For the moment where we have this software that is able to create the AAD specific for me or what I need in that context. In that moment. OK, let's assume this, but this means that the software needs to know who am I, what are my needs and my wishes and so on. So it's profiling me what that means in terms of other parts of society. We're not going to ask for a mortgage or whatever.

1:3:56.107 --> 1:4:2.977 gianmaria.greco@unimc.it Also, that's personalization is very it can be a tool for personalization, but it can also be a tool for.

1:4:3.637 --> 1:4:6.707 gianmaria.greco@unimc.it Ohh, surgical discrimination.

1:4:13.857 --> 1:4:14.987 Kim Steyaert No, thank you very much.

1:4:11.747 --> 1:4:17.17 Alicja Zajdel Thank you. Are there any other questions or comments? Yeah, sorry.

1:4:21.277 --> 1:4:22.607 Alicja Zajdel We are going to say something, Kim.

1:4:23.637 --> 1:4:33.57 Kim Steyaert Hi, yes. I just wanted to say thank you for your answer. Indeed. Uh, do the really things we should take into consideration here. Yeah, but thank you.

1:4:35.717 --> 1:4:36.467 gianmaria.greco@unimc.it OK.

1:4:38.587 --> 1:4:39.217 gianmaria.greco@unimc.it You right now.

1:4:36.7 --> 1:4:39.417 Alicja Zajdel There is another question in the chat. I think then this might be the last one.

1:4:40.17 --> 1:4:49.187 Alicja Zajdel Umm well I and media accessibility. Ohh yeah, media accessibility. I research also sparked a low tech turn.

1:4:51.377 --> 1:4:53.17 Alicja Zajdel OK, I'm going to turn the question is.

1:4:51.417 --> 1:5:8.767 gianmaria.greco@unimc.it Uh, yeah, yeah. I mean, I'll, I'll. That's. Yeah. Not from the apocalyptic, apocalyptic perspective, just that ohrana. I am not so sure if I understood what other perspective there is not apocalyptic if you I I ran.

1:5:9.327 --> 1:5:15.327 gianmaria.greco@unimc.it Ohh, if you can just elaborate on this just so that I can respond on the point.

1:5:15.887 --> 1:5:34.887 Irene Hermosa Ramirez Yes, I was. I was thinking precisely on on this. Well, some people here are working on a green accessibility project and I was thinking on this the turn to no, not artificial intelligence, but more simple solutions, less pollutant solutions. And this is just thinking.

1:5:36.207 --> 1:5:41.17 Irene Hermosa Ramirez Of the resources that artificially intelligent requires.

1:5:42.227 --> 1:5:47.157 gianmaria.greco@unimc.it Ohh well, I honestly I really don't know. I mean that's yes, it's a possibility.

1:5:48.917 --> 1:5:54.447 gianmaria.greco@unimc.it I'm not so sure how much that will be sparked by AI.

1:5:55.837 --> 1:6:9.397 gianmaria.greco@unimc.it Because AI within media accessibility is the even the debate, it's not there, but in a more, you know, widespread. It's not part of our core topics of discussion, but.

1:6:10.77 --> 1:6:21.257 gianmaria.greco@unimc.it Discussion about going low tackle. Let's say as you in that sense it's already there, so it wasn't sparked by AI by that, that it's sparked by the fact that technology is.

1:6:21.337 --> 1:6:24.787 gianmaria.greco@unimc.it So A not environmentally neutral.

1:6:25.947 --> 1:6:31.377 gianmaria.greco@unimc.it And the dates another concern. So I don't think it will be sparked.

1:6:34.607 --> 1:6:41.437 gianmaria.greco@unimc.it But they said that that that I kind of low tackle, low impact on the low tech, low impact.

1:6:43.317 --> 1:6:50.897 gianmaria.greco@unimc.it No, that is helpful. But in everything in the world, probably in every aspect of our life or most of them.

1:6:55.787 --> 1:7:6.157 Alicja Zajdel So I think if there are no other questions, then we should not take up more of your time. But once again, thank you so much. You are leaving us with a lot of food for thoughts.

1:7:5.937 --> 1:7:14.387 gianmaria.greco@unimc.it Uh, OK, Alicia, there was a. Someone gets ania. So can I still ask?

1:7:14.877 --> 1:7:15.727 Alicja Zajdel Go ahead and.

1:7:24.117 --> 1:7:25.897 Alicja Zajdel I don't know if she's going to ask him the.

1:7:24.667 --> 1:7:27.467 gianmaria.greco@unimc.it A Anna ask.

1:7:24.57 --> 1:7:29.7 Anna Jankowska I I will try. I am. Yeah, I'm I am working. Can you hear me?

1:7:29.587 --> 1:7:30.117 Alicja Zajdel Yes.

1:7:29.977 --> 1:7:30.677 Anna Jankowska Yeah, no.

1:7:31.877 --> 1:7:32.157 gianmaria.greco@unimc.it Yeah.

1:7:31.357 --> 1:7:58.57 Anna Jankowska Sorry, I'm I'm walking on the street. I was trying to type, but it's raining cats and dogs and snow, so it's a bit impossible. But I was thinking this or always with the with the AI or whatever it is machine related. There are voices especially in our industry, about taking jobs away and for me always when I look into something that could, you know, maybe improve the workflow.

1:7:58.867 --> 1:8:5.597 Anna Jankowska For the benefit of the the users, but there is always the shadow. OK yeah, but this will take jobs away.

1:8:6.237 --> 1:8:10.147 Anna Jankowska Uh, what do you think about that? And this this question?

1:8:10.827 --> 1:8:14.77 Anna Jankowska In our research, and if we engage in something.

1:8:15.87 --> 1:8:15.897 gianmaria.greco@unimc.it Uh.

1:8:14.847 --> 1:8:16.547 Anna Jankowska What are your views on that?

1:8:17.127 --> 1:8:43.827 gianmaria.greco@unimc.it Thanks, Ania for this is a this is a normal debate, especially if you think about machine translation, right. And Sabian was the conference and the conference that I mentioned before the European Translation Conference, whatever is the name and part of that we should start let's stop AI in general and that specific version of AI which is machine translation was really the.

1:8:43.907 --> 1:8:53.77 gianmaria.greco@unimc.it And the the, the, the, the, the mood of the of the, of the room and I.

1:8:54.187 --> 1:9:5.897 gianmaria.greco@unimc.it OK, let me be honest. I think it's a. A doesn't really make sense. I mean, make sense to socially OK humanly makes sense. You are worried about but the.

1:9:6.647 --> 1:9:18.667 gianmaria.greco@unimc.it The question should be yes, uh machine translation machine translation or and AI are going to are already having an impact on.

1:9:26.627 --> 1:9:27.167 Sabien Hanoulle (Guest) If you.

1:9:27.227 --> 1:9:30.297 gianmaria.greco@unimc.it Of media accessibility experts OK in.

1:9:31.427 --> 1:9:40.117 gianmaria.greco@unimc.it But then it means we need to change. The human will always be in the loop, but precisely because there are aspects.

1:9:40.817 --> 1:9:42.337 gianmaria.greco@unimc.it That are.

1:9:43.977 --> 1:9:53.507 gianmaria.greco@unimc.it We are really good at doing and it's not about to crunching data, OK, it's not about. It's not based on crunching data. It's about being human.

1:9:54.257 --> 1:10:6.117 gianmaria.greco@unimc.it Uh, uh, the I, we will change. We will have a new jobs, we will have, we will change the type of jobs, print the invention of printer has a.

1:10:8.347 --> 1:10:39.517 gianmaria.greco@unimc.it As a A decimated or extinct, the the you know what I mean. The more monks copied or that was a job that was a mission number. What was also a job? Well, now there are editors, proofreaders and so there will be something we need to talk about. What are try to look, what are the jobs of the futures and how to make our our to the use of AI.

1:10:40.337 --> 1:10:44.977 gianmaria.greco@unimc.it To work better, to improve our workflow to work better, to produce better quality.

1:10:45.377 --> 1:10:54.157 gianmaria.greco@unimc.it Uh, a better quality products and have a better quality job instead of of jobs. Instead of let's stop it because.

1:10:54.257 --> 1:10:57.527 gianmaria.greco@unimc.it Or or I'm gonna lose my job. I don't know.

1:10:59.537 --> 1:11:7.387 gianmaria.greco@unimc.it The new jobs will be created, you know, let's let's talk about the what should be the new jobs instead of the jobs we're gonna lose.

1:11:13.237 --> 1:11:14.147 Anna Jankowska Thank you.

1:11:21.147 --> 1:11:24.627 Alicja Zajdel OK, great. So I think we will leave it there.

1:11:25.307 --> 1:11:31.717 Alicja Zajdel Umm, thank you so much. Once again, thank you for everyone for coming and just stay up to date with our events.

1:11:34.457 --> 1:11:35.407 Alicja Zajdel Thank you, Germany.

1:11:34.137 --> 1:11:42.137 gianmaria.greco@unimc.it Thank you so much. Thank you so much a lot for having me. If I may leave you with the one thought.

1:11:43.397 --> 1:11:50.957 gianmaria.greco@unimc.it Just read the reread The Imitation Game and instead of writing I put translation.

1:11:51.667 --> 1:11:53.567 gianmaria.greco@unimc.it Instead of playing chess, put translation.

1:11:54.257 --> 1:12:3.967 gianmaria.greco@unimc.it We thought for a long time that being able to translate was some translator text was something that was inherently human, and they're required to be human.

1:12:4.867 --> 1:12:20.387 gianmaria.greco@unimc.it Now there are things. So what is the role of? What is translation is changing the meaning of translation is at least challenging the meaning of translation. And what is the role of translators? What what defines the translate if the machine is able to translate?

1:12:21.57 --> 1:12:34.137 gianmaria.greco@unimc.it Good. And they are improving better and better than what makes me as a human translator back. Just I'll leave you with this. So think about that because I think it's something that is worth thinking about for each one of us.