Audiovisual Translation: The Road Ahead

The audiovisual translation (AVT) sector has undergone rapid changes in recent years. It would be uncontroversial to state that the various stakeholders: academics; freelancers; technology providers, and language service providers (LSP) are likely to hold diverse and interesting views about what the future holds and how they might be called upon to adapt to recent and future changes. We have conducted qualitative research with representatives of these stakeholders in an attempt to ascertain their concerns and also their predictions for the future. Our motivation was to discover where stakeholders see the sector in the next 10 years. The research was conducted in 2016 and 2018 during the Languages and Media conference on a sample of 160 experts from various sector stakeholder groups. The findings show a broad range of issues that can be summarised into three main themes: the status of the language service provider; the need for standards and metrics; and the importance of training.


Introduction
Research that involves predicting the future is uncommon, since prediction is hard and unreliable. However, it is in the very nature of research to use facts available and predict future trends, and such activity is fascinating and provokes discussion. Whilst the prediction of future tendencies and developments forms the basis of many a conference panel and plenary discussion, it is rare that views and opinions expressed in these and similar forums are collected and evaluated in a systematic way. Our research set out to collect this data using qualitative methods and investigate the beliefs and attitudes of a cross-section of highly qualified professionals in the audiovisual translation sector. Understanding stakeholders' projections and concerns is interesting academically, and is also significant in gaining a picture of the general direction of the audiovisual translation industry to inform future research.
Why predict the future of AVT? Let us go on the ground, or rather under it. Anyone who travelled on the trains of London Underground, the Tube, 20 years ago and has revisited them in 2020 before Covid-19, will notice at least one difference. Apart from the increase in the price of the journey and somewhat more dilapidated trains, something has changed in the habits of Tube travellers. In the past, most users of the Tube who wanted to alleviate the boredom of the repetitive journey by entertaining themselves would read a book or newspapers. There are still readers on the Tube, but in 2020 large numbers of people watch audiovisual content on their smartphones, tablets, or computers. Viewers no longer need to wait for the specific broadcast time of their favourite TV show. They can simply download it, or stream it should network connection be available, and watch it on their smartphone, tablet, or portable computer whenever and wherever they wish. Cisco (2019, p. 2) predicts that globally " [c]onsumer Video-on-Demand (VoD) traffic will nearly double by 2022. The amount of VoD traffic by 2022 will be equivalent to 10 billion DVDs per month." Having worked in audiovisual translation as practitioners, respectively as a freelancer and as an LSP employee, and as academics, our experience tells us that expectations of these various AVT stakeholders may differ. We were interested to see whether these various viewpoints also influence expectations of the future, and whether all stakeholders see the same road ahead. This article will present our raw data and an analysis of this data in this position paper that attempts to predict the future of AVT, or, at least, to reflect the trends that are perceived by those people working in the industry in various roles.

Background
In common with many other areas of human endeavour that involve technology and people, AVT has been changing rapidly. Over the past 20 years or so audiovisual translators have gone from working for public broadcasters using VHS tapes, DOS-powered software, and analogue technology, to working for streaming and video-on-demand services (SVOD) such as Netflix in a cloud-based environment using digital technology (Bywood, Etchegoyhen, & Georgakopoulou, 2017). Even though the world of AVT has always been in constant flux, one could argue that the fast turnaround times for translations became more prevalent and technology started to play a more prominent role when the DVD was introduced: With the appearance of DVDs on the market, developments in the subtitling industry during the last part of the twentieth century have posed a challenge, creating the demand for multilingual subtitling in a large number of languages with very tight turnaround times. The response from the market has come in the form of centralisation, whereby large subtitling companies offer services for all languages required, minimising the costs of Hollywood studios since there is no longer the need to send master copies all over the world. (Georgakopoulou, 2009, p. 31).
A more recent development in broadcasting is the introduction by broadcasters such as the BBC of streaming and catch-up services that enable viewers, for instance, to see the TV series of their choice or a TV show such as Dr Who after, or sometimes even before, their traditional broadcast times and dates. These services are not confined to viewers in the broadcaster's territory, with services such as Britbox, a collaboration between the BBC and ITV, enabling access to content for subscribers worldwide. Netflix and Amazon Prime, another SVOD, are fast encroaching on the territory of traditional broadcasters by producing their own content. Language service providers (LSP) need to handle the production of subtitles, dubbed and voiced content for a wide variety of clients in many languages. On top of that, audio description for the blind and partially sighted, as well as subtitling for the deaf and hard of hearing have become common in many countries. In a nutshell, the amount of work in the audiovisual translation sector has increased significantly over the past 20 years and the nature of that work has changed in various ways (Georgakopoulou, 2018).
In territories where once dubbing was the dominant mode of audiovisual language transfer, such as Spain, SVODs now also offer subtitles, enabling viewers to select their preferred way of watching translated language content. Freelance audiovisual translators are faced with their own set of challenges, such as decreasing fees for their translation and the fact that in most cases they are not the owners of their translated products once they send them to an LSP, which many see as a breach of copyright. Academics are shifting their focus from solely describing processes involved in AVT, to taking a more proactive approach and conducting experimental research, trying to find out more about viewers' preferences for and experiences of watching subtitled or dubbed audiovisual content (Díaz-Cintas & Szarkowska, 2020).
Various stakeholders in the translation industry tend to have different views on how best to improve workflow and working conditions. Technology plays an increasingly important role in the translation industry: "Technology, far from replacing humans, is instead a tool that helps them keep up with surging demand for high-quality translation," (The Economist, 2015, February 5) the journal concludes, optimistically. Freelancers may fear technology may endanger their jobs or make their work unprofitable at the very least, while on the other hand LSPs struggle with fast turnaround times and competition. Our main motivation for this research was to examine the attitudes of those people at the centre of these dilemmas and use this data to attempt to map where the audiovisual translation industry is heading.

Methodology
We gathered qualitative data from audiovisual translation sector stakeholders at the Languages and the Media conference in 2016 and carried out a similar exercise at Languages and the Media in 2018, basing the second session on our findings from the first, in order to test the continuing validity of the findings and examine the state of attitudes and opinions at this second gathering. We used the method of opportunity sampling (Schofield, 1996;Jupp, 2006) and our target population was a subset of the attendees of the Languages and the Media conference, an international conference that takes place in Berlin every second year. Languages and the Media is the longest-running conference in the AVT field and attracts top-level participants from all sectors of this translation field. The 2018 conference report states that of 378 delegates, 66.4% represented business and 21.7% education, training, and research (Languages & the Media, 2018).
Our approach involved self-selection of participants and was not controlled for any particular variable other than attendance of one or both conference sessions, in 2016 and 2018. The rationale for the use of such a sample was that Languages & the Media is the largest and most comprehensive gathering of audiovisual translation stakeholders, a population that would be able to provide rich data in response to our questions. It was therefore our aim to tentatively predict the direction of audiovisual translation in the next ten years by collecting data and opinions from a range of standpoints. The Languages and the Media (L&M) conference series is a uniquely rich environment since it avoids the homogeneity of purely academic conferences and tradeshows where only industry participants gather. It is a conference aimed at all stakeholders in audiovisual translation, including these two groups, but also freelance subtitlers and dubbing translators, technology providers, representatives of professional bodies and the specialist press. Data was collected during two 90-minute sessions, at Languages and the Media 11 in 2016 and Languages and the Media 12 in 2018. In both cases, we prepared a set of questions (Appendix 1) to be answered, in small groups, from the point of view of specific stakeholders. That is, regardless of the actual role embodied by the individual, the group were to attempt to answer the questions as if they were working in the role assigned to them. One role was assigned per group. Although people tended to revert to their own points of view, the device of asking participants to attempt to take another perspective on an issue was intended to create a starting point for debate within the groups; to mitigate some of the extreme opinions sometimes expressed in question and answer sessions at similar conferences; and to make for a more balanced and rich discussion. In this, the strategy was successful. In addition to this, in 2018 we made use of audience interaction technology in the form of the Slido application, 1 which allowed us to ask questions and gather anonymous free-text feedback from participants via their smartphones (Appendix 2). The stakeholders were freelance audiovisual translators, language service providers, content owners, technology providers and academics. The questions were introduced at the beginning of each session, and participants, sitting at circular tables, were asked to take the position of one of the stakeholders and consider a series of targeted questions. A2 paper and marker pens were provided to each table to facilitate data gathering and these written records were collected and transcribed following the session. In both 2016 and 2018 there were eight groups of between 6 and 12 participants, and each group was given 20 minutes to discuss the questions. This was followed by a 40-minute plenary session in which each group reported back to the room the results of their discussion and other comments were welcomed. As each group was reporting, we took notes on a flip-chart visible to all participants in the room. These notes, along with the table notes, the Slido data, and notes taken by the researchers after the sessions, form the basis of our analysis presented in this paper.
In 2016, this exercise was conducted for the first time. In 2018 the session was envisaged as a recap and commentary on the predictions of 2016. In 2016 there were around 100 participants, and in 2018 there were around 60 participants. Only around 20 participants attended both sessions. In 2018, the participants were given an insight into the data collected in 2016. At the later conference, the participants were also asked to evaluate whether the predictions were still accurate two years on.

Findings
In terms of the experimental design, the availability of a large sample of participants was positive in gaining enough data to yield significant conclusions. The device of asking people to answer questions from a particular point of view was useful in removing some of the emotion that often characterises such discussions. However, it was not applied uniformly to the discussions, and so responses should be viewed as general viewpoints, and not representative of a particular stakeholder group. This does not invalidate the data collected but affects the way the data should be regarded.
In light of this, rather than present the findings from each stakeholder perspective, where these might not be representative, we instead focus on significant common trends that emerged from the feedback from the groups and the plenary discussion, highlighting the areas where there was broad consensus between the groups in 2016 and two years later. It should be noted that the main focus of the answers offered by all the groups was on the subtitling part of the sector, reflecting perhaps the make-up of the participants and the main localisation method used by most content owners, in terms of minutes translated. An overview of all issues with the corresponding attitudes expressed by the majority of participants is presented at the end of the section.
There were six areas where there was consensus between the two groups of participants. Both groups, in 2016 and 2018, agreed that: • bigger LSPs are necessary; • there should be better cooperation between academics and industry; • students should be taught transferrable skills; • some skills are assumed by employers; • standardisation and harmonisation are necessary for the successful future of the sector; • technology should be used to rate subtitles.
This section will elaborate on each of these trends in turn, detailing the main and salient points of the discussion. The discussion section will then draw together these themes and examine whether the opinions and predictions can be shown to have any veracity in the intervening time since the 2018 conference.

Status of the Language Services Provider
In 2016, we were interested to see what participants felt was the outlook for the following 10 years for a Language Services Provider. More specifically, our question was: "What will a translation agency look like in 2026?" Views expressed included the opinion that LSPs make excessive profits compared with other actors in the supply chain; that working conditions need to be more flexible, including schedules and places of work; and that LSPs need to be more proactive in protecting the copyright of individual subtitlers. Overall a very strong trend in both conference sessions was that LSPs needed to increase in size.
Since the amount of audiovisual content is growing, and content owners include not only television broadcasters and film distributors, but also e-learning providers and SVOD platforms, the prediction was in 2016 that larger LSPs would be necessary to keep up with increasing volumes. This view was confirmed in 2018, with participants expressing the opinions that LSPs need to both increase in size in terms of personnel, but also have teams that cover more roles and functions within each standalone entity.
In both sessions there were lively discussions about the future of the LSP and the likely developments in the relationship between freelancers and the said companies. In 2016, the majority view was that freelancers would work directly with end clients, rather than through LSPs. This view was partially motivated by the then newly introduced Hermes, Netflix's bespoke platform for testing potential subtitlers and monitoring the quality of their. At the time, Netflix appeared to be setting up their structure such that they could work directly with freelance subtitlers, and this was certainly the belief of a number of our participants. Netflix closed the platform in 2017 less than a year after its launch. According to Allison Smith, Globalization Innovation Manager, formerly Program Manager in Globalization Strategy and Implementation, at Netflix, the reasons for discontinuing Hermes are as follows: Netflix aimed to own the full process from subtitler recruitment through to working in our tools, this started with Hermes. While a lot was learned and there was value in the test, after introspection and analysing our core competencies, we decided that vendors were better suited to use their core competencies and add value to the content localization ecosystem by owning the recruiting, training and onboarding processes. (Smith, personal email communication, December 2018).
In 2018, the participants in the session were still of the opinion that freelancers were likely to work with clients directly more frequently, taking the view that new technological solutions will enable them to do so more and more.
As a perhaps natural consequence of this finding, there was also a prevalent opinion in the 2016 session that the LSP as a concept could be significantly endangered, and some heated discussion resulted. This view was probably motivated by the discussion around the introduction of the Hermes Test. Smith's comment above also shows Netflix seriously intended to work directly with freelancers and thereby remove the need to interface with any LSPs. Two years down the road, in 2018, the participants of our workshop no longer were of the opinion that LSPs were going to disappear, but instead took the view that they would transform, and that they would focus more on technological solutions. This was the narrative feedback gathered. However, we also asked participants in 2018 a specific question via Slido, namely: "Is the LSP as we know it obsolete?". The findings were interesting and are presented in Figure 1.

Slido Data: Is the LSP as We Know it Obsolete?
Note: data taken at the conference, see Appendix 2 As is apparent, there is no evidence to suggest that the LSP is obsolete, as 49% of the 37 participants who responded answered in the negative. However, it does seem that a large proportion of our respondents, some 43%, believe that the LSP as we know it is endangered.

There Should be Better Cooperation Between Academics and Industry
We also wanted to know what challenges academics (translation trainers) will face over the next ten years, including how the student profile will change.
The idea of cooperation between academia and LSPs has existed for some time, and the Languages and the Media conference series is a product of that idea, since one of the aims of the conference is to bring together AVT professionals from all corners of the sector to enable networking, knowledgesharing, and collaboration. Many universities, for instance, successfully cooperate with LSPs on various activities, including teaching, research, and authentic projects. However, participants in the 2016 session thought that this cooperation was insufficient, and that deeper and better cooperation was necessary. This view was also confirmed in 2018, although participants did not clearly specify how that should be done or what form it might take.
The idea that academia should respond to the needs of the marketplace and technological advancement is not a new one. It is often propagated by governments who feel that this close cooperation could improve employment, and hence strengthen the economy. As the providers of AVT training programmes, it would seem incumbent on the universities to be agile in meeting the needs of the industry. However, some participants of our session in 2016 felt that academia finds it difficult to keep up with the pace of change in technology, since investing in new technology requires financing and university curricula can be difficult to amend. To a large part, this view is supported by the nature of academia in many countries. In higher education, processes can take a long time for various reasons, such as applications for grants, and long-winded approval and decision-making processes. However, it should also be noted that academia also plays a role in supporting the industry through research that can inform practice and process.

Students Should be Taught Transferrable Skills
This widespread opinion is linked with the previous point. In both sessions, our participants felt that transferable skills, including project management, should be taught at university level, since they are necessary in the market. There was a view that teaching project management in translation at university level can easily enable students to develop skills that are needed for project management in other sectors or generally in the translation industry when working as a freelancer.

Some Skills are Assumed by Employers
Certain technical skills such as computer literacy are now generally assumed by employers, according to our participants from 2016. Computer literacy was indeed not assumed 20 or so years ago and there was no expectation that students be taught skills such as project management. This seems an uncontroversial finding that was also confirmed in 2018. It was felt that it was the role of universities to prepare students of translation for the world of employment by teaching them such skills as would be necessary for them to undertake entry-level jobs or work as freelancers. This links directly with the previous point and perhaps calls for more precise delineation of the particular skills that employers require, something we return to in the discussion section.

Standardisation and Harmonisation are Necessary for the Successful Future of the Sector
A strong trend in the answers from the first session was that without some kind of industry quality standardisation, in terms of both translation and technical standards, it would be impossible for audiovisual translators to protect themselves from price dumping and aggressive discounting by unqualified practitioners and LSPs that are lacking in subtitling experience and qualifications. This quality standardisation might include both translation and technical elements, but specific details of how this standardisation might be achieved were not elaborated upon in the time available. In 2018, the view of the participants was that this process of standardisation and quality evaluation was not happening rapidly enough, and that technology could be leveraged to aid with this, as in the recommendation below.

Technology Should be Used to Rate Subtitles
This claim from 2016 is connected to the option SVOD providers offer to their users, rating content. One idea of our participants was that similar rating systems can be used to rate subtitles, so that content owners have an idea whether their chosen suppliers are delivering the level of quality they desire for their end users. In 2018 this idea was generally supported by our participants, but there was a concern that this would only work if it were done correctly and in a manner that is fair to the subtitlers. There remains some suspicion of the end client and their motives in this respect. A related idea was that small royalties, termed "nanopayments" (Samid, 2015, pp. 53-56), could be arranged for subtitlers for the usage of their subtitle files. For instance, a subtitler would get a small payment from Netflix or a similar broadcaster each time the programme with their subtitles is streamed. However, this raises an issue of copyright, something that was also debated at the session but will not be discussed here for reasons of space. This idea was popular in both groups of participants, even though there was no indication that there is a move in that direction on the part of the end clients.

Overview of Findings
The table below presents an overview of all the topics discussed and the attitudes displayed during the sessions in 2016 and 2018. Table 1.

Discussion
With a rich and varied a sample of around 160 expert participants, it is possible to ascertain some trends and go some way towards testing the temperature of the sector within the framework of the questions asked. Both groups of participants, in 2016 and 2018, included people from all stakeholder groups in the industry. It can be assumed, given their professional profiles, that they were familiar with the most prevalent issues in the industry and brought their own personal viewpoint to the table along with that of the role they were asked to embody. Qualitative research of this nature is valuable for showing trends in thinking and opinions within a group of experts in a particular field and thereby indicating the prevailing attitudes to sector developments. In what follows, we examine the overall tendencies demonstrated in our data, and match these against developments since the last conference in 2018, in an attempt to test the validity of the tentative predictions and assertations made and to begin to assess the efficacy of such research in the audiovisual context.

Industry Consolidation and the "Death" of the LSP
The most striking change in opinion was concerned with the shift in the view of the future of LSPs in just two years. While our 2016 participants had been of the opinion that LSPs were doomed to extinction, mostly influenced by Netflix's aim of working more with freelancers directly through its Hermes platform, in only two years the view had changed, since the platform proved to be unsustainable for Netflix, as confirmed by Allison Smith (see section 4.1). LSPs are not extinct although there is still the belief that they are endangered and perhaps the view that they need to change, adapt, and consolidate to deal with the changes in the wider industry context. There was a noticeable view that as clients become bigger, LSPs need to correspondingly increase in size and in the diversity of services offered, most probably through consolidation, that is merger and acquisition. This viewpoint from both the 2016 and the 2018 sessions has been reflected in developments not only in audiovisual translation but also in the wider translation industry since the data was collected. Georgakopoulou (forthcoming, p. 10) discusses the strong trend for increasing mergers and acquisitions in the AVT space; Bond reports 60 mergers and acquisitions in the LSP space in 2019 in contrast with 48 deals in 2018 and states "M&A was also the key driver of growth for many industry Leaders" (Bond, 2020). Research specialists Nimdzi also agree with this assessment, commenting on the growth of the language industry in 2018 "[t]he increase largely comes from mergers and acquisitions" (Nimdzi Insights, 2019) and in 2019 "We can expect the M&A [mergers and acquisitions -emphasis added] wave to continue in 2020" stating that their research indicated that almost 46% of LSPs are planning either to acquire companies in the near future or are actively looking for potential buyers" (Nimdzi Insights, 2020).
What can also be concluded from our data is that stakeholders consider Netflix to be the most important player in the industry. As demonstrated in the word cloud in Figure 2, "Netflixification" has been coined as a neologism, and even though there are other SVOD players and traditional broadcasters offering similar services, such as Disney +; Amazon Prime; HBO Max and Peacock TV, what Netflix chooses to do is, or at least was in 2018, perceived to have a significant effect on the rest of the sector.

Need for Standards and Metrics
Participants in both groups strongly expressed that without some form of agreed quality standards covering both technical and translation-related quality, the supplier side of the sector is vulnerable to price dumping and the ingress of unqualified practitioners. There was an overall sense that the profession of subtitling would benefit from harmonisation in terms of the assessment of quality, for both intralingual and interlingual subtitling. Quality standards do exist, of course: broadcasters and content owners such as the BBC and Netflix have guidelines and quality documents, and the NER model (Romero-Fresco & Martínez, 2015) is used in several settings for quality assessment of intralingual live subtitles (e.g. Ofcom, 2014). Pedersen (2017) gives a comprehensive overview of the various approaches to quality assessment for both intralingual and interlingual subtitles, and suggests his own model for quality assessment of interlingual subtitles: the FAR model, which focuses on assessing Functional Equivalence, Acceptability (adherence to target language norms) and Readability. He acknowledges several weaknesses in this model, including acknowledging that it may "seem time-consuming or complicated" (2017, p. 224) and that there is a large amount of seemingly inevitable subjectivity in the application of the criteria. To date, there are no published applications of this model in a commercial setting, and many broadcasters and content-owners who do carry out quality control seem to rely on a combination of the "second pair of eyes" model with a set of guidelines for this process. Indeed, as Georgakopoulou (forthcoming, p. 16) asserts, "quality will continue to be a hot topic in the modern translation pipeline."

Training and Talent
Many of the viewpoints expressed during the data gathering focused on the need for attention to be paid to the training of new subtitlers and AVT professionals and also the need to address the gap between what LSPs desire from their freelancers and the skills and competences of new graduates. This perceived discrepancy between what universities are teaching and what the industry requires is a common topic of discussion where diverse stakeholders are gathered and there have been various initiatives proposed to address this including the current AVT working group of the European Master's in Translation 2 and the Elia Exchange programme 3 (aimed at the whole translation industry and currently suspended) as well as specific smaller collaborations between industry partners and universities. The shortage of trained professionals for subtitling and related jobs has been reported by Estopace (2017), who cites increased content volume thanks to streaming services; and Howorth (2019), speaking from an LSP perspective, amongst others. The solution to this talent crisis is elusive. Programmes that teach solely AVT are expensive to run and difficult to set up at university level in a sector that suffers increasingly from the requirement to be profitable. Clearly this issue deserves more research.
It is interesting to note that we could see from the answers to individual questions that most responses and issues were negative. For instance, many participants expressed their concern regarding the influence of machine translation on the remuneration of translators. This is despite all the questions being phrased in a neutral manner and the discussion being moderated by both academics without bias as much as possible. However, we asked the participants at the end of the 2018 session, using Slido, whether they were more or less hopeful about the industry than they were in 2016. Participants' responses were overwhelmingly positive, with 51% of respondents reporting they were more positive, and 24% reporting they felt the same. Note: the data taken from the conference, see Appendix 2 Our final piece of data-gathering was the request, via Slido, to submit a word or phrase that summed up how the participants felt about the industry and its prospects over the next five years, the following word cloud was generated, indicating again that the majority of the submitted words were broadly positive.

Word Cloud Indicating How Participants Felt About the Prospects for the Industry Over the Next Five Years
Note: data taken at the conference, see Appendix 2

Conclusion
As discussed, the audiovisual translation sector is no stranger to change and adaptation, and currently finds itself in the midst of another such transition period. However, there are numerous examples of such changing working environments around us. Radio did not destroy newspapers, and newspapers have survived and adapted to the internet. TV did not destroy either radio or newspapers; DVD did not destroy TV, and it seems SVODs will not destroy TV or cinema, though the 2020 pandemic may force some radical changes in this area.
Our research aim was to crystallise the issues and the concerns of a wide range of stakeholders from within the audiovisual translation industry in the hope of ascertaining the focal points of concern for the years to come. The data gathered indicates that the three themes discussed above-namely the role and status of the language service provider in the changing environment of audiovisual translation, the need for standards of quality in terms of both translation and also technical aspects, and the desire for well-trained talent to serve the needs of an ever-expanding market-are likely to frame the development of the audiovisual translation industry for the years to come.
A future experiment might set up the groups so that they only contained stakeholders from a particular part of the industry, although this might weaken the discussion or lead to more bias. The relationship between broadcasters, LSPs and freelancers could further be investigated, especially changing workflows due to the arrival of machine translation. The relationship between the words that came up in small and large print would also merit further examination through similar qualitative data gathering. Feedback from participants anecdotally was overwhelmingly positive and there was certainly value in conducting such a session in the conference context. The use of Slido in 2018 provided a more formal way of data gathering, and this could certainly be extended in any future study. Although prediction is most certainly hard and unreliable, we have shown that qualitative research such as this can yield useful and insightful information that, in this case, does indeed give an indication of future directions.
• Copyright is out of LSP control.
• Working conditions need to be more flexible.
Have the predictions been correct so far and do they still pertain? What are your predictions with respect to this particular question for the next two years? Any general predictions to add?

Questions to consider FROM THE POINT OF VIEW OF AN (avt) ACADEMIC
Two years ago we asked you to consider (from an academic perspective): What challenges will academics (translation trainers) face in the future?
How will the student profile change over the next 10 years?
You said: • Academia and LSPs should work together.
• Pace of change in the technology world difficult to respond to in a university context. • Transferable skills should be taught.
• Certain skills (IT) are now assumed.
• Project management should be taught more.
• Students should be given a realistic overview of the industry and working conditions. • Students should be taught to demand appropriate remuneration for their work.
Have the predictions been correct so far and do they still pertain? What are your predictions with respect to this particular question for the next two years? Any general predictions to add?

Questions to consider
Two years ago we asked you to consider the following two questions.