Watch this YouTube and read the file attached then start answer the questions.
Digital Immortality and Virtual Humans
Maggi Savin-Baden1 & David Burden2
Published online: 27 September 2018 # The Author(s) 2018
Abstract The use of virtual assistants such as Siri that provide voice and conversational interfaces, the growth of machine learning techniques to mine large data sets and the rise in the level of autonomy being given to computer-controlled systems all represent shifts in artificial intelligence that are enhancing the creation of digital immortality. The growth of personality capture and levels of brain simulation as well as computationally inspired life after death may change the future of religion, affect understandings of the afterlife and increase the influence of the dead surviving in society. This paper provides an overview of recent developments in the area of digital immortality, explores how such digital immortals might be created and raises challenging issues. It presents the early findings from a study that created a virtual persona. This prototype system contains relevant memories, knowledge, processes and modelling of an individual’s personality traits, knowl- edge and experience and, also, incorporates the individual’s subjective and possibly flawed view of reality. It is argued that this system offers the possibility for the development of a persona that learns post-death.
Keywords Digitalimmortality.Virtualhumans.Chatbots.Autonomousagents.Artificial intelligence . Machine learning
This paper presents research on digital persona creation and reflects on the possibilities for the development of a persona that learns post-death. The idea of creating digital immortality is raising concerns from those in the field of software development. In a
Postdigital Science and Education (2019) 1:87–103 https://doi.org/10.1007/s42438-018-0007-6
* Maggi Savin-Baden [email protected]
David Burden [email protected]
1 University of Worcester, Worcester, UK 2 Daden Limited, Birmingham, UK
recent study, Maciel and Pereira (2013) found that religious and moral values affected the perception of death, which in turn affected design solutions. Studies such as this illustrate how cultural, political and religious beliefs can affect the technical landscapes around the design of digital immortality. Companies such as Eter9, Lifenaut and Eternime are all purporting to be offering chatbot-based versions of digital immortality, but the challenges of creating an effective digital immortal are considered here within the context of creating a virtual persona of a living person for use whilst they are still alive.
Research and Literature
The possibilities for creating digital immortality have become more sophisticat- ed through technological advances. For example, Google Duplex was recently able to book a restaurant table verbally without the other party realising that they were talking to a computer (Leviathan 2018). There are companies that are dedicated to creating digitally immortal personas such as Eter9 (Bearne 2016), Lifenaut (2017) and Eternime (2017) and the now defunct Intellitar (Fussell 2016Facebook has now put in place measures to control the post-mortem data on their site (Brubaker and Callison-Burch 2016), although this does not appear to be as effective as many people would wish. Steinhart (2014) has examined personality capture, mind uploading, and levels of simulation, arguing for a computationally inspired theory of life after death that will change the future of religion radically.
The idea of being able to live on beyond your natural death has a long history in our culture and remains popular in novels, such as The Night’s Dawn Trilogy by Peter Hamilton. Prior to our technological age, the agency for this was typically the ghost, and the twentieth century saw examples from The Ghost and Mrs Muir to The Sixth Sense. In the digital era, most of the Artificial Intelligences within science fiction have tended to be ‘evolved’ artificial intelligence, such as SkyNet in Terminator and Ultron in the Marvel films, which have become sentient rather than being created as digital immortal personas of other people. However, there are examples, particularly in the science fiction of the last decade or so, where the artificial intelligence are digital immortals, or at least are the digital copies of real people. As well as Caprica (the Battlestar Galactica prequel, where Zoe Greystone has been creating a copy of herself, but she then dies leaving the digital copy to carry on her legacy), there is also Eunice, a digital copy of a now dead matriarch from Alisdair Reynold’s On the Steel Breeze (Reynolds 2013) and Poseidon’s Wake (Reynolds 2015). The potential of digital copies of dead people is tackled head-on in the ‘Be Right Back’ (Brooker 2013) episode of the Black Mirror TV series. A slightly different take is found in the novel Kiss Me First (Moggach 2014) where a computer-savvy loner ‘covers’ for a woman who commits suicide, extending her electronic presence beyond her death. What is interesting about all these construc- tions is that the digital persona is very much living in the here and now of their progenitor’s death, rather than facing up to the implications of potential immortal- ity, which is reflected in the rise of digital endurance concepts.
88 Postdigital Science and Education (2019) 1:87–103
Digital Endurance Concepts
Digital endurance concepts have developed in order to make sense of the ways in which digital technology is being harnessed to commemorate and memorialise the dead. These can provide useful frameworks in which to consider digital immortality issues:
Kasket (forthcoming) argues that online persistence and the ongoing presence of the data of the dead online will lead to more of a globalised, secularised ancestor veneration culture, and it is important to recognise the ongoing persistence of the dead online on social media, LinkedIn, Amazon, and YouTube.
The Restless Dead
Nansen et al. (2015) take this further and argue that forms of digital commemoration are resulting in cultural shifts towards a restless posthumous existence. Thus, there is a shift away from the idea of death as being sleep or rest (Hallam and Hockey 2001), towards the restless dead, as they materialise through social media and technical capabilities. Such media include living headstones, digitally augmented coffins and commemorative urns embodying the head of the deceased, and which are seen to interrupt the previous limitations of cemeteries, static headstones and biological death.
One-Way Immortality and Two-Way Immortality
Savin-Baden et al. (2017) suggest that ultimately, the important distinction will be between forms of digital grief which are essentially one-way (or passive) and those which are two-way (or active). The former is where the recipient can read about the deceased in some form of digital memorial, either intentionally created (for example, using a site such as SocialEmbers web service or the KeepTheirMemoryAlive mobile phone app) or an existing system which lives on after their death, such as a person’s in- life profile that has been put into memorialised/remembering status in services such as Facebook. Two-way immortality is where there is the possibility of the digital entity interacting with users and visitors, and with the rest of the living world, in the form of a chatbot or virtual human. This interaction could be in a wide variety of forms, from two- way text or even voice and video conversations, or by the virtual human manipulating real world systems such bank and investment accounts (Godfrey 2018; Dzieza n.d.).
Digital Grief Practices
Digital grief practices are particular practices that have developed through media and digital media and that have become acceptable norms and are the most common ways in which digital grief is likely to be encountered today. Although less sophisticated than the virtual personas to be discussed later, they do show how some of the initial concepts of digital grief are entering into our culture:
Postdigital Science and Education (2019) 1:87–103 89
Media mourning is defined here as the idea that we are urged to mourn something that is not our grief through social media, such as the 2017 Manchester Arena bombing, or to mourn our personal loss through social media in a highly public way.
Walter (1996) argues that the purpose of grieving is to construct a durable biography that allows survivors to continue to integrate the deceased person into their lives and to find a stable and secure place for them. In practice, this now tends to occur more often through digital memorisation and the use of companies such as Eternime.
This is the process of memorialising people through avatars in online games and 3D virtual worlds. An example of this is ancestor veneration avatars (Bainbridge 2013) that are used as a medium for memorialising the dead and exploiting their wisdom by experiencing a virtual world as they might have done so.
Digital commemoration is a practice that crosses the boundaries of digital immortality and digital legacy to provide particular commemoration services. These include memorials and tribute pages hosted on special memorial sites; ceremonies, such as funeral and memorial services in 3D Virtual Worlds such as Second Life; solar power headstones with a Quick Response (QR) Code (the matrix barcode) that provide information about the deceased; and digitally mediated funeral practices, such as augmented coffins. There are also free services to enable you to light a candle, such as the Centre for Healing Arts and the Walser Funeral home where you can pay for a memorial candle, monthly, annually or forever. Death management company Everplans provides a list of typical digital commemoration services at https://www.everplans. com/articles/the-top-10-online-memorial-websites.
Some people wish to take a more proactive role in managing their post-death presence. For example, websites such as Dead Social enable users to instruct their Facebook and Twitter accounts to post-future updates after they are dead, such as pre-prepared birthday messages. It prompts you to create a social media will, quoting the UK Law Society, presumably as a means of self-legitimation. Whilst the service is free, Dead Social has distinct enrolment periods and during enrolment, they allow 10,000 users to subscribe to the service. It is also increasingly common for people to create a digital will in which they indicate what is to be done with their digital legacy and assets, and includes, passwords and security questions.
90 Postdigital Science and Education (2019) 1:87–103
With the emergence of such digital immortalities, it is useful to identify the different people who are likely to encounter such digital legacies, including relatives, friends, lawyers, politicians and religious leaders. These may be of three principal types:
& Preservers—who use memories and artefacts to create a legacy, such as a repre- sentative avatar or digitally immortal persona, and which may include the deceased themselves before death
& Receivers—who receive the memories and artefacts, including representative ava- tars or digitally immortal persona
& Mediators—professionals who encounter legacies, representative avatars or digital- ly immortal persona, such as priests or lawyers
It is clear that there needs to be an understanding of what seems to be the emergence of the different ways that ‘creating’ digital immortality can both result from and create digital legacies and traces and the impact these may have on both recipients and wider society (Harbinja 2017).
Three potential types of creators of digital immortalisation can be identified:
& Memory creators—those creating passive digital memories and artefacts pre- and post- the subject’s death. These have already been considered above in examples such as virtual venerations, digital commemoration, digital memorisation and durable biographies and are typically not created by the subject.
& Avatar creators—those creating a representative interactive avatar pre-death, typi- cally by the subject, which is able to conduct a limited conversation with others but has a very limited capability to learn, grow, act on and influence the wider world around it (and hence could be considered a virtual humanoid in the typology identified by Burden and Savin-Baden 2019). It has minimal likelihood of being mistaken for a still-living subject.
& Persona creators—those creating a digitally immortal persona pre-death that learns and adapts over time and can influence and act on the wider world around it (and hence could be considered a virtual human or even ultimately a virtual sapien (Burden and Savin-Baden; 2019). It has a high likelihood of being mistaken for a still-living subject.
All of these are likely to draw on a person’s existing digital legacy. The digital legacy comprises any information that exists in digital form after death and includes social media profiles, email, online shopping accounts, digital music and photos, as well as account information, digital assets, and digital property, things that are static once the user has died (Bassett 2015, 2017). The growth of machine learning techniques to mine large data sets such as these, making deductions from them that can equal human analysis, and the rise in the autonomy being given to computer-controlled systems is all having a relatively unknown impact on society when the original subjects are living, as well as when the subjects are deceased.
Increasingly, these digital legacies are in the form of ‘digital traces’—the digital footprints left behind by interaction with digital media. These tend to be of two types:
Postdigital Science and Education (2019) 1:87–103 91
intentional digital traces—emails, texts, blog posts, Facebook and photographs—and unintentional digital traces—records of website searches, logs of movements and phone calls. Accidental residents of the digital afterlife who leave unintentional traces are seen as internet ghosts or the ‘restless dead’ (Nansen et al. 2015). The somewhat eerie consequences and impact on recipients are unclear, particularly in relation to ancestor veneration avatars where people are immortalised as avatars in online role-playing games (Bainbridge 2013). The traces may be intentional creations pre-death, or unin- tentional for the dead but intentional by those left behind.
Digital Immortality Creation Options
There are already companies that are actively trying to create computer applications which are predicated on the creation of avatar-level digital immortalisation. It can be hard to separate rhetoric from reality, but a survey of the three such systems Eter9 (2017), Lifenaut (2017) and Eternime (2017) is provided below:
Eter9 describes itself as ‘is a social network that relies on Artificial Intelligence as a central element’ and that ‘Even in your absence, the virtual beings will publish, comment and interact with you intelligently.’ A key element are Counterparts:
your Virtual Self that will stay in the system and interact with the world just like you would if you were present. Your Counterpart will learn more with each action you take. The more you interact in the new social network, the more your Counterpart will learn!
Such a Counterpart is able to continue to post and interact with others on the network after you are dead (Morse 2015). In practice users, post tweet style messages (‘thinking into Eternity’) which can be read by other users of the system (if set to ‘public’ or ‘connections’), and also, it is assumed, start to build the knowledge base of the Counterpart (although how this happens is not detailed). There are also ‘Eternilisations’, effectively favourite posts. Eliza Nine, the host bot has (as at 3 Jul 18) 54467 connections, which is probably a reasonable estimate of the user-base to date, active users seem significantly less than that. There is no obvious way in which you can access your own or other’s Counterparts in order to see how well they are developing, if at all.
Lifenaut works on a similar principle to Eter9, enabling people to create mind files by uploading pictures, videos and documents to a digital archive, but this is an explicit process rather than a background one as with Eter9. It also enables the user to create a photo-based avatar of the person that will speak for them, although there is the choice of only a single male and female voice which are both US English. Lifenaut is a product of the Terasem Movement Foundation (http://www.terasemmovementfoundation.com/)
92 Postdigital Science and Education (2019) 1:87–103
which describes its work on Lifenaut as being to investigate the Terasem Hypotheses which state that:
1. a conscious analogue of a person may be created by combining sufficiently detailed data about the person (a ‘mindfile’) using future consciousness software (‘mindware’), and
2. that such a conscious analogue can be downloaded into a biological or nanotech- nological body to provide life experiences comparable to those of a typically birthed human.
In practice, Lifenaut offers a number of ways to build a ‘mindfile’. These include:
& Just talking to a few sample bots, although the conversations seem non-sensical and they appear to ignore what you say
& Filling out some interview questionnaires, including some ‘validated’ personality profiles, including a 486 question personality survey measuring cautiousness, conscientiousness, cooperation, gregariousness and nurturance.
& Talking to your own avatar so it learns from what you say, although this appears to require an explicit ‘correction’ action
& Manually adding favourite things and URLs, although your bot does not appear to learn them once added.
It is however also possible to have your mindfiles beamed continually into space for later, potential, interception and re-creation by alien intelligences. It is notable that much of the site functionality is implemented using the now deprecated Adobe Flash and so hard to access from modern web-browsers and completely unsupported by the end of 2020 (Mackie 2017).
When using Eternime, the individual is expected to train their immortal prior to death through daily interactions. Data are apparently mined from Facebook, Fitbit, Twitter, e-mail, photos, video, and location information with the individ- ual’s personality being developed through algorithms through pattern matching and data mining. Eternime is currently in a private stage (as at June 2018, and at July 2018), so it is not yet possible to verify any claims or really understand what technology is in use. As of 3 July 2018, 40.497 people were signed up for their wait list.
Given the practical state of Eter9 and Lifenaut, and the continued ‘private Alpha’ of Eternime, it would seem that there is more hype than substance to much of the current media coverage of such digital immortalisation systems. There are also other high- profile android/chatbot projects which should be approached with caution. For exam- ple, in Autumn 2017 Sophia, a humanoid robot gave a ‘speech’ at the United Nations (United Nations 2017) and has even been granted citizenship of Saudi Arabia (Griffin 2017). Whilst this may encourage recognition of the fact that there needs to be more debate, as well as legislation in this area, the level of technical accomplishment actually shown is hotly debated (Ghosh 2018).
Postdigital Science and Education (2019) 1:87–103 93
Creating a Virtual Persona
In an attempt to understand what is and is not currently possible in terms of creating a chatbot which could form the basis of an avatar-level digital immortalisation, Daden, KSharp, the University of Worcester and the University of Warwick1 have been working to create a virtual persona, which is a digital representation of some of the memories, knowledge, experiences, personality and opinions of a specific living physical human (Savin-Baden and Burden 2018). This work should be contrasted with, for example, the digital copies of Holocaust survivors Pinchas Gutter and Eva Schloss created by the USC Shoah Foundation (McMullan 2016) which work off a fixed store of around 1000 recordings of sentence level responses and have no flexibility in what they do. The intention, in the current study, is that a user would be able to interact with the Virtual Persona in the same way as they would with its physical subject, and that the Virtual Persona would present the same highly subjective and possibly flawed and biased information, views and opinions as its physical subject.
The scenario that is driving the development of the Virtual Persona is not the creation of a digital immortal per se, but rather the creation of a persona which could be ‘left behind’ when an employee moves from one job to another. It could be accessed by their successor in order to get advice on how to do the job, and opinions and information on projects, clients, customers, suppliers, technologies, procedures and staff. The project and persona are very much seen in the context of knowledge management and intended to explore how a virtual Persona could aid in knowledge capture, retention, distribution, access and use.
In the context of digital immortality though, the ability to access the acquired knowledge, experiences and insights of someone who is no longer available could offer guidance and support for those still living. Indeed a digital immortal would be created almost by accident if a persona were created for the purpose of retaining corporate knowledge if the subject suddenly died prematurely in an accident or from a terminal disease. It is the likelihood of such unfortunate incidents occurring that require an early consideration of the ethical, moral and legal issues around digital immortality.
For the purpose of the experiment and in order to keep within ethical guidelines, the subject was allowed to both filter and edit material provided where they felt that the real information could be compromising, since in reality, they were still continuing in their job. It should also be noted that creating a chatbot of a specific person being ‘interviewed’ in a one-on-one situation where the user/interviewer knows that they are talking to a chatbot places a very high bar on any attempt to pass the Turing Test— the benchmark for evaluating chatbots (Turing 1950). In contrast, the authors have been involved in building chatbots for two ‘covert’ Turing Tests, where the participants did not know they were talking to a chatbot and where the chatbot represented a generic personality and, in these cases, has achieved deception rates (i.e. percentage of users thinking they were talking to a human) of 80% (Gilbert and Forney 2015) and 100% (Burden et al. 2016). Thus, whilst being able to completely fool a user was beyond the scope of this project, there may well be lower levels of performance which can be
1 Sponsored by Defence Science and Technology Laboratory (Dstl)—in collaboration with Dstl Technical Partners
94 Postdigital Science and Education (2019) 1:87–103
achieved which still yield a useful tool, and the knowledge and expertise gained along the way may also have application in other areas of knowledge management. Since the virtual persona was based on a subject called Barry, it was only natural that it became referred to as Virtual Barry.
Virtual Barry Development
Virtual Barry was developed in an iterative manner over an approximate a 2-year period. Several reviews were held to check progress and identify areas for further work, which included:
& face-to-face interviews followed by audio transcription & Skype voice interviews & Skype text-chat interviews & answering questions loaded onto an interview application so that the Subject could
answer questions asynchronously & completion of spreadsheet grids to ensure consistent data collection on topics such
as customers, projects and employees & import of curated data from the Subjects’ mobile phone, web browser, address book
and calendar & entity extraction on social media posts and documents produced by the Subject
A disadvantage of several of these were that (a) the Subject was aware of the human involved in collecting the information and (b) the data had to be coded manually. In the later stages of the project, the Subject was given the ability to chat with the Virtual Persona. Throughout 2018, qualitative and quantitative user evaluations have been undertaken by people who either knew the Subject or knew the Subjects type of job and areas of expertise, or who knew both. This data is currently being analysed and will be the subject of further papers. This paper will present a broader reflection of some of the lessons learnt in virtual persona development and then examine the implications for digital immortalisation.
An image of the current (April 18) Virtual Barry interface is at Fig. 1.
Virtual Barry Reflections
During the Virtual Barry project, it became evident that the virtual persona was manifest through a combination of different elements within the system, which included the user interface, content, word choice and conversational style. The project is also looking at a more direct mapping of psychological profile information onto Virtual Barry, but that will not be considered here. However, the idea of there being different ‘flavours’ of the persona have emerged. Each of these elements will be considered in turn.
If the virtual persona looks like Barry and sounds like Barry, then users are probably less inclined to focus on the content of what Virtual Barry says. Considering these two elements:
Postdigital Science and Education (2019) 1:87–103 95
Whilst the Sitepal system has enabled the team to animate an image of Barry, it loses some believability as it speaks, with the only just acceptable lipsync, ‘white space’ for teeth and facial distortions during some expressions. Having a better animated avatar is likely to improve the believability of Virtual Barry. As noted previously, the user interface is not a high priority for the project and the Sitepal system is still relatively unique in terms of ease of use and integration. Better systems are available as bespoke applications or within 3D tools such as Unity3D, but would take more time and effort to integrate. The project team continues to seek a better solution.
Until recently, only a relatively limited number of voice fonts were available from the main text-to-speech companies which has severely limited the ability to match Barry’s voice (for example, although Sitepal provides access to 4 leading text-to-speech engines covering almost 100 voices they offer, only 3 British English male voices and 5 British English female voices). However, in 2018, there has been a growth in companies offering personalised voice fonts, such as LyreBird (https://lyrebird.ai/), VocalID (https://www. vocalid.co/) and Modeltalker (https://www.modeltalker.org/). The team is evaluating these to see if they offer a credible alternative within the constraints of the project. In addition, Google has also released its own text-to-speech system, which whilst not offering a better match, does seem to provide a more modulated and expressive generic voice.
As well as factual and interest information, Virtual Barry also needs to reflect the practical expertise of Barry. At the moment, Virtual Barry has very little of this expertise
Fig. 1 Virtual Barry interface
96 Postdigital Science and Education (2019) 1:87–103
111>) 0 ()
select an Emoji or type a response
User: [email protected]
SO, I can answer questions about KSharp, our staff, processes, policies, methods and technologies. I can also help you make some of the typical decisions that face me in the job, such as hiring, purchasing and bidding.
Q Q Q: Q Q @ ~ay,-thiag ————————–
0201a. Daden Limited
knowledge (in Human Factors in the Barry case), and this needs to be added to the system. However, much is procedural and related to decision choices and processes and, so, forms more of an aspect of procedural memory, which may be less amenable to the knowledge- graph approach and reflect instead aspects of more traditional expert systems.
Word and Phrase Choice
Verbatim responses from Barry which are encoded into Virtual Barry naturally have a high degree of ‘Barryness’. However, the longer such responses are, the fewer situa- tions in which they will work. There is also a discontinuity when Virtual Barry moves from a constructed response to a verbatim response and back again. A key goal has to be to try and minimise that discontinuity and enable ‘Barryness’ to show through in shorter and more machine-generated responses. One way of minimising this disconti- nuity is to ensure that constructed responses are using the phrases and words that Barry would use. This two-stage process requires ensuring the responses sound human rather than machine like, and then tuning them to Barry’s vocabulary and style. It is also notable that there is a difference in even Barry’s ‘voice’ between text conversations and spoken conversations, and so the team has switched to using text conversations via Skype instead of audio interviews in order to collect data that are a better match.
Beyond Virtual Barry—Creating Digital Immortality?
Virtual Barry illustrates the two