“Seances are very similar to conversations with chatbots. In both cases, you cannot see the entity that has an unimaginable amount of data. Plus, you can’t always trust the information they give you,” he tells Hypertextual Sean Scanlan, Director of Creative Technology at Deeplocal, a company that has just launched Séance Bot, a ouija board with artificial intelligence.
Based in Pittsburgh, in the northeast of the American map, this company focuses on the development of interactive experiences that “serve to tell stories.” As revealed, the team includes engineers from NASA, Disney, the Department of Defense and universities. They describe themselves as “creative engineers”.
“We started as an offshoot of an art and technology laboratory that operated at Carnegie Mellon University. Shortly after, we were at the forefront of the first use of robotics in advertising,” says Alie Cooligde, communications director of the group. Her reference is to Nike Chalkbot, a campaign carried out more than a decade ago, in which an automaton wrote inspiring messages on the asphalt tape, intended for Tour de France cyclists. Now, the Ouija board with artificial intelligence seeks to achieve visibility of that and other company actions.
How Séance Bot works, the Ouija board with artificial intelligence
“Our works bridge the physical and digital worlds. In the case of Ouija, we were interested in including artificial intelligence, a central theme this year,” says Scanlan. The gadget they developed combines a series of technologies and charms. The user sits in front of the table, asks a question out loud, a program recognizes the query, and without human intervention—at least direct—the answers appear on the table. Here, the magician’s secret: the little board moves with a system of magnets. For the rest, these 21st century séances They involve terrifying, albeit fictitious, characters. In that sense, the proposal focuses more on technological creativity than on ghostly contacts.
How did the idea to create a Ouija board with artificial intelligence come about?
Usually, our experiences with artificial intelligence make them feel magical. Previously, we created a portal that transforms people into cartoons, and a installation with flowers that open in response to the movements of those visiting Google’s facilities in New York. There are many themes and stories that involve magic and motivate us to work. The idea of creating the Ouija came from thinking that it is always believed that another person is moving the planchette, the table used to mark letters and form answers. We thought: what would happen if it moved on its own?
We know that Séance Bot works with a massive language model (LLM). Can you tell us more about it?
We have performed custom systems training for different projects. For this particular one, we are left with a standard solution, Google PaLM 2. Since we wanted to communicate with popular figures in the public domain, there was little that an overly advanced model could add to the experience. This allowed us to focus on engineering to get interesting and actionable answers. Also to dedicate more time to the mechanics of the board and the user experience.
Which characters can be contacted using this Ouija board with artificial intelligence?
As I pointed out, they all belong to fictional plots. On the list are the monster of the Black Lagoon, Frankenstein’s Monster, Dracula, the Mummy, the Invisible Man and the Wolf Man.
Can you tell us about projects you are working on?
We are particularly excited about the trend toward experiential retail, where technology and creativity can transform shopping. In addition, we see great potential in incorporating playful and interactive elements to hotel spaces, conferences and trade fairs. Finally, we believe the auto showroom experience is ready for reinvention and we’re working on it.
As we noted, Deeplocal’s AI Ouija board is aimed more at entertainment than at dark science enthusiasts. However, the advancement of these technologies has allowed—and is expected to deepen—certain experiences for bonding with deceased people. At first glance, it seems crazy. However, there are precedents. For example, Mark Zuckerberg himself mentioned that possibility.
In a interview recently with podcaster Lex Fridman, Meta’s CEO said that metaverse users could interact with dead people. For greater rigor, he commented that the conversations would be with virtual avatars loaded with information from those people. “Is that a possibility you really think about?” Fridman asked. The Facebook founder responded: “There may probably be ways to relive certain memories. It could be helpful for those who lost a loved one and are grieving.” Later, he acknowledged that this technology could be harmful to mental health. “I’m not an expert on this. I think we should study it and understand it in detail,” he concluded.
In 2020, before the brutal success of ChatGPT, generative artificial intelligence and the brand new Ouija with artificial intelligence; A unique campaign in the United States used a variant of AI, deepfakes or deepfakes, to “revive” a young man who was killed in a shooting. A video showed him advocating disarmament. The boy in front of the camera was not precisely him, but an emulator created with those technologies. On the occasion, his parents recognized that this was a way to “make someone speak that no one can ignore,” and they mentioned that it has been an “extremely difficult experience.”
As we pointed out previously in Hypertextual, chatbots to interact with the afterlife are not exclusive to the series Black Mirrorwhich masterfully narrates that possibility in the chapter I’ll be back right now, from the second season. Beyond the screens, a well-known case in the field of “technological conversations with the dead” is that of the company Replika. It was created by a woman who lost a friend in 2015, and she developed a program that emulates interactions using audio and voice messages that she compiled in a database. This system is called “Romanbot”, in homage to the late Roman Mazurenko, and is part of the so-called “griefbots“, which we can translate as “robots of grief or mourning.”
A study of 2020 published by researchers from the Open University of Catalonia, in collaboration with Aalborg University, in Denmark, addressed this intersection. The central questions: is it ethical to use automated systems, such as griefbots, to recover the deceased, even virtually? Ouija-like systems with artificial intelligence, could they hinder the necessary grieving processes?
Belén Jiménez Alonso, a doctor in psychology and lead author of the study, said that these systems “would fulfill a psychological function, allowing people to interact in a more sophisticated way with the memory of the deceased person.” The specialist added: “It is important to open a social debate about grief and death (…) In our experiences, the technology of the moment has always intervened, since before it was done in an analogical way, through writing and of letters. Now, social networks or chatbots are new technologies that are transforming the way we deal with it,” she noted.
Scanlan, from the company that has just presented its Ouija board with artificial intelligence, believes that “there are many ethical and moral questions about the use of AI,” and points out that these concerns are especially important if the technology will be used to “communicate.” ” with real people who died. “Because of these questions and ambiguities, we deliberately chose fictional characters to keep this light and entertaining. However, there are some fascinating use cases that are starting to be addressed, for how we can use artificial intelligence to somehow contact the dead. We look forward to seeing how the technology develops and how these questions are answered,” she closes.