Lifenaut Project. Eternalizing Consciousness using a virtual avatar

A talk with Nick Mayer, Manager of Cyberbiological Systems at Terasem Movement Foundation.

The goal of the Lifenaut Project is to create an automated, user-friendly, and web accessible system that allows individuals around the globe to create significantly detailed personality archives.

The service is completely free and our hope is that future AI-like software will allow these archives (which we call "mindfiles") to replicate the consciousness of the individuals from which they came, explained Nick Mayer, whose job is to guide the development teams involved in this project and oversee the orchestration of their work.

Presently the mindfiles are composed of: an array of file types that the user uploads (and can view) using our lifenaut interface, personality test results, profile information, and lines of conversation from teaching sessions between the user and his/her talking "AI Avatar." Mayer believes that although at this point the AI Avatars are not too intelligent, ultimately all of the information in the mindfile can be tapped into the conversational capability of the avatar, thus creating an analog of the users’ consciousness.

More details about the Lifenaut project are described in the following excerpt from a letter that Mayer composed to lifenaut userbase:

The purpose of this note is to provide some context for Lifenaut.com and to introduce some other ongoing cyber-biological research efforts. The Lifenaut Project is a research experiment. Its long-term goal is to test this hypothesis (an educated guess): Given a comprehensive database, saturated with the most salient aspects of an individual’s personality, future intelligent software will be able to replicate this individual’s consciousness.

Why are we embarking on such a project?
The wisdom of ancestors has been valued by most cultures since the beginning of humankind. The Lifenaut Project will not only ensure that ancestral wisdom will continue to guide future generations, but will exponentially improve this connection between past and present by allowing children to have conversations with the actual words of their great, great grandparents, for example.

What if this experiment doesn’t work?
If, after creating a database saturated with the most pertinent aspects of an individual’s personality (what we call a mindfile), future intelligent software is NOT able to replicate this individual’s consciousness, then all of the hard work that individual put into creating their mindfile will not have been wasted. I say this because all of the information about that individual— thousands of lines of free flowing and guided conversations, personality test results, the array of uploaded file types articulately tagged with descriptions, dates, locations etc. etc. will now have been stored in a non-degradable (digital) form. This unique digital archive could potentially exist forever and be accessible to people for generations and generations to come.

How can you say that a “mindfile” could last forever?
The Lifenaut.com database is backed up daily in 3 different locations around the world and in 2 different media types. As new digital storage media develop with greater longevity and stability, lifenaut.com will adopt these new technologies. The key point is that the form of storage is digital. Digital media can be copied and transferred an infinte number of times and will not degrade. Paper degrades, so do cds, photos, dvds, hard drives, videotapes, computers, iPods. But digital information transferred from one medium to another does not. Microsoft actually has a research project underway for which they have filed a patent, entitled The Immortal Information Storage and Access Platform. Something like this could potentially be the final destination of our Lifenaut database.

I am personally excited about the prospect of storing the bulk of my life’s unique information in an immortal database. I view it as an interactive time capsule where my children and even their children’s children will be able to login and learn from my experiences, hear stories exactly as I told them, long after I am gone. It is absurd that we spend our lives experiencing and learning about the world and then at the very end when we have attained the most educated, informed perspective we can possibly have, it all vanishes. Within a few generations all that is left of that lifetime of learning is possibly a name remembered, maybe a story, possibly a few mysterious photographs.
The Lifenaut Project and research towards immortal data storage platforms can reverse this paradox.

Who is conducting this experiment and what are some other projects like it?
While at first this may seem more like sci-fi than science, all one needs to do is read about some of the mind- blowing leaps and bounds that are being accomplished by researchers in the fields of Cognitive Computing/ Artificial Intelligence, Nanotechnology, Neuroscience, and Robotics (resources provided at Science Supporting Terasem). Of particular interest are three research projects whose ultimate goal is to replicate the human brain and its learning process: Numenta, IBM’s Blue Brain Project, and Novamente.

Numenta is a new for profit company spun- off from the Redwood Neuroscience Institute both of which are located in Silicon Valley, CA, USA. Both entities were co-founded by Jeff Hawkins, author of On Intelligence (a short read I highly recommend) and CEO of Palm Computing. Numenta’s approach to developing intelligent bio-software is a fascinating and unique one—they are modeling software after the human neocortex’s actual structure. Like an infant’s set of freshly interconnected neurons, their hierarchical arrangement of “nodes” are learning to visually recognize more and more complex objects. Jeff calls his model Hierarchical Temporal Memory.

The Blue Brain Project is a collaborative effort between IBM and Ecole Polytechnique Federale Da Lausanne. Proffesor Henry Markam has been directing the project since its start in 2005 and research is primarily being conducted in Lausanne, Switzerland. Like Numenta, they are using the human cortex as a model, but their approach is more of a hardware- based one. A press release last month reported that they had successfully modeled in structure and function a cortical microcolumn. One could consider a microcolumn the basic functional unit of the brain—the next level up in organization after a neuron. This was an endeavor that took over 2 years. 30 million synapses down, less than a trillion to go!

Novamente LLC is another project striving to create thinking computers by modeling biological systems. Spearheaded by Artificial General Intelligence author/ researcher Ben Goertzal and futurist writer Bruce Klein, their team is attempting to replicate the infant brain with a learning protocol based upon Piaget’s Stages of Cognitive Development. Phase I of the project culminated in the modeling of a dog’s learning process. They are teaching their cyberpuppy in virtual environments and unveiled their new AI engine at the International Virtual Worlds Conference last fall.

Like these other research projects the Lifenaut Project exists in this nebulous overlap between biology and technology, however it differs from those above in seeking to replicate neurological memory, not processing. Over the past two years author/ entrepreneur Dr. Martine Rothblatt, Bruce Duncan, and myself have worked together with three software development teams to make the Lifenaut Project’s first two phases become a reality. These teams are: Connect Studios, based in RI, USA; iCogno Ltd. Of Dawlish, UK; and InterMediaLab of Norwich, UK.

Modeled after the organization of human cortical memory, the Lifenaut memory storage system is both hierarchical and intra-relational in nature. It is for this reason that the requirement exists for all files to be so richly tagged. Our biggest challenge has been the fact that we are dealing with a diverse array of data that trigger the formation of unique neural networks, but we are not storing the actual neural networks themselves. However as neural headsets become more perceptive and more affordable, we plan to integrate EEG associations with your AI Avatar conversational data as well as with uploaded files/ memories. Two neural headset companies I have been speaking with are Neurosky and Emotiv. The progress they have made over the past year is truly amazing. In addition to the info in uploaded files, presently the database consists of validated personality test results, profile data, and all conversational associations generated by your teaching sessions with your personal AI Avatar.

The short-term (5 year) goal of the Lifenaut Project is to provide a free, engaging, entertaining, and socially interactive environment in which thousands of people like yourself have saturated your personal databases, your mindfiles. For most people we expect that this process will take years. As time goes by we will continue to offer more and more ways for your avatar to become more and more like you.

So have fun crafting your time capsule for future generations and help us guide humanity in a positive direction by building a mindfile.

Regards,

Nick Mayer
Manager of Cyberbiological Systems
Terasem Movement Foundation