Virtual Children, Real Problems

Another Frontier of AI that Rejects Bodily Limits

As the digital revolution continues colonizing space, time, and humanity, another new frontier is on the horizon: virtual children. That’s right, digital babies—cared for by real people. Lest you think this is just science fiction, watch this TEDx Talk about Baby X:
 

Coupled with the large language models being used in AI development (see Robin Phillips’s recent article on ChatGPT-3), it is quite possible to envision a society where digital children are understood as sentient beings with feelings, rights, and independent thought. One such social robot, Sophia, was already granted citizenship in Saudi Arabia back in 2017 (publicity stunt or not).

Expert in human-computer interaction, Catriona Campbell, lays out a series of possible predictions in her new book AI By Design: A Plan for Living with Artificial Intelligence. She suggests that if couples choose to “have” virtual children over real children in even just a fraction of cases, it could help in achieving net-zero energy, alleviate the climate crisis, and reduce overpopulation.

But this rationale fails on its face. Raising children in the metaverse might be done virtually, but the components needed to do so are real. Digital living incurs massive physical costs—like the mind-boggling electricity consumption of the server farms that make the internet possible. Not only that, when declining marriage and birth rates are considered, the population challenge for societies in the long run might actually be population decline.

But most of all, the idea of virtual children offers a sad commentary on life in the modern West. It reveals a lack of convivial hope and is yet another distorted attempt to imitate the natural gifts of embodied life in a way that only increases our false sense of autonomy and control over nature and each other.

Soul Machines
Mark Sagar, CEO of Soul Machines is one of the driving forces behind Baby X, which takes a “holistic approach to embodied cognition.” Sagar explains, “our philosophy is that a virtual brain requires a human-like embodiment to fit in a social world it can interact with, in order to learn and express in general human-like ways.” This is what Sagar calls “experiential learning,” where “Baby X proactively learns through direct experience with people and content. Through interactively playing with the world around her, she discovers things about it in the way that we do.”

While the idea that the brain requires an embodiment is a step in the right direction, the mind-body unity that is the human person can never be replicated with man-made means. Baby X’s “brain” may be “embodied” on a screen or even in robotic form, but Baby X’s connections between body and “brain” are still determined by the designer and thus are accidental in nature, meaning they are not the essence of the thing, but are determined by other causes or means.

There is also a hidden circular argument in Sagar’s explanation—as in much of the AI world. This is something philosopher Hubert Dreyfus has been critiquing for decades. It goes something like this: scientists and researchers utilize a machine or computer analogy to describe how the brain works. Then, scientists and researchers make a computer or AI robot that functions like that model of how the brain works. Then, we say, “it must be intelligent since it functions like how (we assume​) our brains works.”

Informatizing the World
Another aspect of AI that is problematic is the reduction of human intelligence to information. As computing improves, we can collect data and information about anything and everything, which is what Baby X is always doing—just like smart phones, smart watches, smart appliances, ChatGPT-3, and more. Philosopher Byung-Chul Han calls this “informatizing” the world, that is, the measuring and recording of everything—data, algorithms, and analytics rule the day. Even objects become receptacles of data, and so it is with Baby X. Baby X is not a person; it is an information terminal that constantly collects data and responds based on inputs and programming. Yes, it is very complex and impressive, much more so than Amazon Alexa, or a smart refrigerator which tells you when to buy milk, but it is not categorically different.

This informatizing and data-ist approach to the world leads Han to warn that “we are headed towards a trans-human and post-human age in which human life will be a pure exchange of information.…The future of humans seems mapped out: humans will abolish themselves in order to posit themselves as the absolute.”

Han also offers an intriguing response to the growing claims of artificial intelligence that I think are applicable to Baby X and the idea of virtual children. Simply put, Han says, “artificial intelligence is incapable of thinking, for the very reason that it cannot get goosebumps. It lacks the affective-analogue dimension, the capacity to be emotionally affected, which lies beyond the reach of data and information.” In other words, the fact that the mind affects the body and vice versa makes human intelligence unique. Even though Baby X can “show emotion” and respond to human interaction in a variety of ways, it does not feel emotion in its body. The plastic of the computer does not break out in goosebumps spontaneously, nor does the screen cry tears. Even if an AI robot could get goosebumps or cry tears, it would be based on the programming of a designer who built that capability into the machine, based on certain stimuli and response signals, not because the robot actually felt moved.

Limits and Bodies
Returning to Han, the body plays an inimitable role in what it means to be human in community. Han argues that “without bodily touch, no ties can emerge,” and that “community has a bodily dimension.” But, “because of its lack of corporeality, digital communication weakens community.” Han focuses in specifically on the importance of eye contact—what he calls, the gaze. “The gaze stabilizes community,” and “the absence of the gaze is partly responsible for the loss of empathy in the digital age….The gaze of the mother, in particular, provides an infant with stability, self-affirmation, and community. The gaze builds primordial trust. Without the gaze, a disturbed relationship to self and others develops.” Imagine how disturbed the relationship to self and others would be with virtual children, with which there can be no true gaze.

While virtual children may or may not really catch on, the following truth remains. With virtual children, you remain in control. But you are never in control of a real human baby, as any parent will tell you. They place demands on you, they reveal the limits and finitude of embodied life together. They are gift and burden, all in one—and all wrapped up in one body.