- Microsoft has recently been granted a controversial patent to recreate a deceased person as a conversational bot using their online data.
- The concept of “bringing dead back to life” using technology is not cutting-edge as it may sound at first and it’s already been promoted in the past, generating contrasting reactions.
- New technological advances bring both new possibilities and controversies.
- Despite the risks and uncertainties coming from the possibility of “carrying on as a machine”, it can indeed represent a positive change if well regulated
“Imagine being sentient but not alive. Seeing and even knowing, but not alive. Just looking out. Recognizing but not being alive. A person can die and still go on.” (Philip K. Dick, 1977)
December 1, 2020 is a day we might get to remember in the future. At the very end of one already peculiar year, Microsoft was granted a patent to create a conversational bot from pre-existing social data belonging to a specific person. Digitalization earned its special spot in human lives: but what about its place in death?
In a not-so-distant future, images and voice recordings will be eligible for use to recreate a likewise digital reincarnation of deceased individuals, as well as their private texts, messages, chats, e-mails and letters. Not to mention every kind of online data, such as social media posts. Keeping in mind the boundless volume of information people leave behind during their virtual endeavors, it’s not hard to understand how enough data to imitate a real person’s personality can actually be gathered.
Learning from recurring patterns in pre-existing data is not just a machine prerogative: that’s what we do too. Human reactions and decisions are mostly driven by unaware associations about previously acquired information. What makes this very same process different in artificial intelligence is that they are seemingly faster and more accurate.
In the specific case of the Microsoft patent, a personal portfolio built from online material (such as conversations, pictures and virtual content) can be used to train the bot to communicate according to the user’s tone, manners, conversational attributes and — ultimately — personality, while voice recordings will be put into use to generate a realistic voice font.
However controversial, this is not the first time artificial intelligence is used to bring the dead back. The controversial South-Korean documentary “Meeting You”, aired in February 2020, disclosed the experience of the first woman to use VR technology to see for the last time her long-gone child. The same desire to reunite with a departed relative also moved a young man, back in 2016, to create Dadbot: an interactive chat-bot programmed to communicate like his father would.
Similar controversial enterprises of course elicited various reactions. However, among the many viewers of Meeting you, many expressed their support to the South Korean woman and affirmed they wished they had the privilege to meet for the last time someone they lost. According to responses like this, it could be argued that — despite its apparently questionable nature — the chatbot proposed by Microsoft would not be perceived as creepy as it sounds.
Utilizing sensitive data belonging to a deceased user to reincarnate him as a bot brings several ethical implications to the table: one of them is a matter of privacy. Each scenario involving data acquisition and treatment requires specific data protection laws, this is a matter of fact.
Unfortunately, there is still no specific legal frame that covers the possibilities disclosed by the Microsoft patent. In order to avoid any privacy violation or data exploitation, further regulations will be needed to avoid any post-mortem exploitation of personal information, therefore requiring some kind of code to regulate disclosure permission both from the user and his relative.
A different dimension to consider is the relationship between the bot and the user. If identity can be considered as the persistence of specific attributes in time and the stability of certain first-person memories, would it be therefore correct to affirm that we will be able to transfer our identity to a machine after death? Whatever the answer, it would be easy to confuse the act of chatting with a bot that resembles a person with the feeling of chatting with a bot that replaced that same person.
If this is the case, it should be an active work to generate awareness about the aim, the capacities and the limitations of a similar product before its release, so that consumers can make proper informed decisions, grounded on appropriate expectations.
Lastly, the possibility of being reincarnated in a chat-bot concerns the delicate challenge of grieving. Moving on after the death of a loving one is an emotional trial, during which coming to term with loss is required to reach acceptance. But what will happen to such acceptance in a world when people will be able to make contact with their deceased relatives and friends as much as — or even more — they do with their living ones?
About this regard, it’s still not clear if a conversational AI built on a lost one would impact negatively on the process of grieving; therefore, objections have been made on the idea that the use of technology in this peculiar field would only exploit the pain and suffering of vulnerable victims for the sake of profit.
These sort of ethical dilemmas still remain unsolved for one particular reason: however created from artificial matters, it’s a human job to answer similar questions. Robots are no more a distant concept from futuristic imaginary, but an expanding reality. Even so, focusing on the possible negative uses it’s a common attitude: decades of science fiction educated the human kind to fear artificial intelligence, leading to a biased point of view that prevents many from considering possible applications. Despite this, it’s clear that each impactful technological advancement in human history was paired with the challenge of changing what human beings think and feel about their lives.
The possibility of virtually reuniting with a lost relative may not thrill everyone, but it could be useful to find closure after sudden circumstances that made it impossible to say goodbye to a loved one. One of the toughest lessons of COVID-19 emergency was the breath-taking sense of impotence of being distant from our dearest relatives during their last moments. Conversational chat-bots similar to the ones proposed by Microsoft might at least be a solution to comfort the ones who are going through this difficult path.
On another hand, the possibility to speak to their living relatives may become something that people could express in their will. It’s already common to provide dispositions about the ways we want our memories to be carried on, the fear and pain of leaving them behind. In this case the sensitive data would be utilized only after previous authorization from the willing person and the final actualization of the bot might be authorized post-mortem by the living relatives, not as a desire to replace the person they lost, but as a celebration of their previous life.
Ultimately, agreeing to upload traces of their own lives could become part of a voluntary sign-up system. Donating our bodies to science as a way to provide a better future for the ones we leave behind is a concept we are accustomed to, as well as donate our organs to save someone else’s life. In the very same one, we could be granted in the future the choice to donate our own minds.
Doctor Who, a hugely popular TV show for over 50 years now, utilized this concept in one of the episodes called Silence In The Library part 1 and part 2, in which a computer is built to allow a child’s mind to live on. Sounds familiar?
Future people, according to this episode, can donate their face upon death just like any other organ. These faces can then materialize, say, in a library as a part of a good customer service, when a computer chooses which face you would like best when talking to a help droid.
The concept of keeping our memory alive through a conversational bot isn’t exactly new, but the recent Microsoft patent made it even more concrete. The presence of large amounts of online data makes this possibility even more concrete. The age of digital afterlife is upon us and, despite it leading to many unanswered questions, there is one particular thing we must ask ourselves.
Are we ready for it?
Cramphorn, S. (2006). Blink: The Power of Thinking without Thinking / Strangers to Ourselves: Discovering the Adaptive Unconscious. Journal of Advertising Research, 46(1), 135 LP — 138.
Kübler-Ross, E., & Kessler, D. (2005). On grief and grieving: Finding the meaning of grief through the five stages of loss. Simon and Schuster.
Mouammine, Yahia & Azdimousa, Hassan. (2019). Using Neuromarketing and AI to collect and analyse consumer’s emotion: Literature review and perspectives.
Ruane, Elayne & Birhane, Abeba & Ventresque, Anthony. (2019). Conversational AI: Social and Ethical Considerations.