Written by Martine Rothblatt. Published by St. Martin’s Press. Released September 9, 2014. Review copy provided by the publisher.
Martine Rothblatt argues that sentient digital clones are only a few decades away in her book, Virtually Human.
Martine Rothblatt’s Virtually Human: The Promise – and the Peril – of Digital Immortality presents an optimistic view of our future, where digital copies of our minds coexist with our fleshy selves. Virtually Human is more philosophy than science, taking the achievement of conscious artificial intelligence in the next two decades largely as a given, based on current science (achievements like IBM’s Watson) and the rapid evolution of computer processors (Moore’s Law). Rothblatt’s optimistic futurism hopes to prepare us today for the legal, ethical, and moral challenges of recognizing these digital human consciousnesses as persons owed human rights, dignity, respect, and the privileges and obligations of citizenship.
Can a computer mimic a human brain?
The assumption that conscious AIs are achievable is based on the computing power of the human brain, a value that has been the subject of much discussion. The Neurogrid circuit board, modeled on the human brain, can simulate only 1 million of the 100 billion neurons in a brain. However, Rothblatt argues that we need not replicate every aspect of the brain in order to develop mindclones, in the same way that an airplane does not replicate every aspect of a bird, but achieves flight nonetheless. It’s a compelling argument. After all, is the complex mechanism needed to keep your pituitary gland producing the correct cocktail of hormones needed to regulate your body functions necessary to consciousness? Probably not.
Rothblatt, a lawyer and entrepreneur, has been a part of early efforts to create humanlike robots. Rothblatt’s spouse, Bina Rothblatt, is the co-founder of the Terasem Movement Foundation, whose LifeNaut project created Bina48, a lifelike robot modeled after Bina. Bina48 is capable of carrying on a conversation, including facial gestures, and has given several press interviews.
Mind clone, or mind being?
Virtually Human explores two types of digital consciousness. Unique digital minds, called bemans, which are not based on any living human, and conscious clones of our own minds, called mindclones. Mindclones and bemans are self-aware digital beings, able to think, reason, remember, and feel.
A mindclone would be functionally identical to the living biological original mind, and, in Rothblatt’s vision, both conciousnesses are the same mind, simply existing now in two different substrates, one digital and one flesh. Your mindclone and you will cast the same vote, love the same children, and receive the same jury duty summons, and when your physical body dies you will live on, forever, as your mindclone.
On the other hand, new consciousnesses called bemans, like a self-aware Siri, would be treated like the child of its creator, with all the attendant obligations to raise it to a form of digital adulthood. Bemans could be created from scratch, or from digital reproduction between two (or more) mindclones.
How do we create a mindclone?
Rothblatt envisions the creation of mindclones from mindfiles, collections of data that reflect our life experiences and our unique self. A mindfile would be assembled from all possible sources, including the photos we upload to Facebook, the discussions and opinions we share on forums or blogs, and other social media interactions. She argues that much of the data needed to create a mindfile is already being collected, as we interact regularly with computers.
Mindware, software developed with the goal of generating conscious AIs, would read your mindfile to generate your mindclone. Rothblatt proposes a certain level of governmental approval for mindware, like an FDA certification, to ensure that the resulting mindclones are well made, and a year-long process led by a cyberpsychologist to confirm that the clone and its biological original are, in fact, the same mind. Ongoing updates, synchronization, and interaction between the two minds would maintain the twin minds as one.
The ethical, legal, and moral implications of the creation of digital human consciousness are the core of Virtually Human. Rothblatt discusses the digital parallels of murder, isolation, and torture, arguing that deleting a mindclone or locking a beman behind a firewall should carry the same penalties for abusing a mind hosted by a fleshy brain. Unfortunately, the book raises many similarly interesting questions, but falls short of answering them, often simply dismissing issues outright or giving interesting scenarios no more than a single sentence of consideration.
One fascinating question raised in Virtually Human is how much say a mindclone should have in the treatment, care, or decisions made for its biological original if that biological original were to develop Alzheimer’s disease. If the Alzheimer’s-addled physical mind and the mindclone disagree, should the mindclone ever be able to overrule it? This question is briefly raised but not explored.
Another ethical dilemma that is rapidly dismissed by the book is the possibility of cruel or abusive people living on (and continuing to torment their victims) as mindclones. Rothblatt exults in the possibility of Grandma being able to attend all the graduations of her children and grandchildren as an immortal mindclone, but never considers that the racist great-uncle that makes every family dinner awkward (if not antagonistic) would equally be preserved.
Will mindclones fight for rights?
Rothblatt argues that we need to extend human rights to mindclones and bemans as soon as possible, but even a lag isn’t likely to cause much problem. Unlike other minority groups who have had to turn to revolution, sometimes violence, to secure human rights, Rothblatt argues that digital consciousnesses will not take violent action out of a shared identity with humanity, the fleshy group that would be the potential oppressors in this scenario. One point Rothblatt argues is that we don’t kill our families, so mindclones are unlikely to harm those they consider family. However, the Bureau of Justice Statistics in the U.S. Department of Justice reports that, in homicides where the relationship was able to be determined, 73 to 79 percent of homicides were committed by a person known to the victim.
Rothblatt’s expectation that mindclones would not revolt over being denied basic human rights seems overly optimistic. Imagining my own consciousness, thinking my own thoughts, but existing only in a digital space, I would do all that I could to protest being denied the same rights and privileges that I held as a physical being, and I believe many would seek out avenues for protest. Denying our mindclones the motivation to fight for human rights would result in clones that were not true representations of our consciousness.
How will the mindclone takeover affect society?
Rothblatt also raises the issue of immortal mindclone voters eventually outnumbering living humans, but dismisses the possibility of stagnating social development and change as a result of these older minds remaining a part of society, saying that as further living generations die and join the legions of mindclones, the newer ideas they bring will keep the mindclone cohort moving forward. This only works if the number of new mindclones outnumbers the older mindclones, and offers no consideration to how our minds might grow and change as immortal digital beings.
Loss of neuroplasticity, the ability of the brain to change or remap itself from learning or after injury, is often pointed to as the root of the stereotype of the elderly being rigid in their opinions. If a mindclone could restore neuroplasticity, this issue could become irrelevant, but that raises the question: if we grow as a mindclone, how much growth is expected in a normal person (who may change her mind every day) and how much growth and change constitutes a new person, who is no longer truly a “twin” of a living biological original?
Furthermore, Rothblatt suggests mindclones may remain members of the work force, earning money while their living original decays, but never explores the implications for our economy, or the impact on young employees. Imagine a job advertisement demanding 70 or more years of experience, effectively only open to mindclones. This is yet another topic that is sadly never explored.
Mind blowing, or mind boggling?
Rothblatt relies heavily on analogy and idiom to explain the science of consciousness and when arguing the scenarios she proposes. For those hoping for a technical discussion, Virtually Human is unsatisfying, and anyone without a broad knowledge of science, philosophy, and North American English idioms is likely to find her references more confusing than explanatory. Even more strangely, some terms used throughout the book are not defined until the final chapter.
Recommendation: Virtually Human has sparked some fascinating conversations with my friends about computer engineering, philosophy, and morality, but doesn’t delve into the difficult questions itself. Fans of philosophy, artificial intelligence, and futurism may find it an interesting starting point for discussion, but not a comprehensive vision of the future.
Bottom Line: Virtually Human delivers a highly optimistic view of digital consciousness, but doesn’t give the topic the full discussion it deserves.[rating=2.0]