Login
You're viewing the oslo.town public feed.
  • Rob Ricciricci@discuss.systems
    Oct 20, 2026, 4:02 PM

    In 2015, Shannon Vallor wrote a paper about technology and moral de-skilling; it is even more relevant now than it was then. This is my take on her argument, extrapolated to 2026.

    * Technological changes (as well as social ones) modify the way we spend our time and the types of activities we engage in
    * This creates a process of upskilling and deskilling, where we (individually, and as a society) gain new sets of skills while others atrophy - this happens both intentionally, and as an inevitable effect of how we now spend our time
    * Acting in accordance with our morals is a skill - it requires intention, practice at recognizing the "right" actions in ambiguous situations, experience to predict outcomes of our choices, and sometimes self-restraint to act against our immediate self-interest in service of outcomes that align with our morals
    * The more we use machines and automation for tasks that have a moral dimension, the more we engage in a process of moral deskilling of ourselves

    Evidence of many types of deskilling are all around us; skills that used to be common because of their necessity, like knitting, woodworking, hunting, food foraging, etc. are relegated to the margins, have become the realm of specialists, or are treated as hobbies, because we have mostly automated away the need for large segments of the population to engage in them.

    Vallor uses examples of warfare (offloading decisions about killing to machines) and caregiving (offloading physical, mental, and emotional support to robots, chatbots, etc.) as cases where we are likely to morally deskill our society; these were somewhat hypothetical in 2015, but it seems clear that we're well along this path in 2026. I think the judicial system, law enforcement, employment decisions, and allocation of resources such as healthcare, education, and social services would also make good examples.The more we automate or outsource these activities, the less good we get at performing them in ways that align with our morals.

    While it's easy to think of this in the context of actions that have a clear, classical connection to morality, it operates in much smaller, and more pervasive ways as well. The way we act towards each other - whether we respect each others' dignity, autonomy, time, and humanity. The way we use common, finite, and shared resources. The decisions, large and small, we make about our interactions with the natural environment. Every one of these, and more, has a moral dimension, and the more we put machines between us and the practices of deciding and acting, the more we erode our ability to make moral decisions and to act morally.

    We consider people who are skilled at acting morally to be virtuous, and those who are not as having poor moral character. As individuals and as a society, our current technological choices are morally deskilling us, reducing our moral character.

    Maybe we should re-consider our relationship with some of this technology, lest we end up in a future where moral reasoning is something that is just done as a hobby, idly, while sitting in a boring lecture, like knitting.

    bhaven.org/uploads/3/4/0/3/340

    💬 2🔄 8⭐ 0

Replies

  • Rob Ricciricci@discuss.systems
    Oct 20, 2026, 4:03 PM

    Many thanks to @viveksrikumar and Thi for making this paper required reading for their AI Ethics class, and to @jbigham for discussions about this yesterday and the vivid knitting analogy

    💬 0🔄 0⭐ 0
  • Oliver Kennedyokennedy@discuss.systems
    Oct 20, 2026, 4:56 PM

    @ricci There's an essay by Larry Niven called "Why Men Fight Wars and What You Can Do About It" (1984) that looks at a narrow slice of the same question. The essay observes that the surrender reflex is necessary (at the species level) as a sort of social de-escalation of last resort. As soon as one party becomes aware that they are clearly outmatched, instinct kicks in and they surrender.

    More than that though: There is not only an instinctual pressure to surrender, but also an instinctual pressure for the aggressor to accept their surrender. This latter bit creates a path (even if a bad one) for tense, intractable situations to de-escalate, and averts self-sustaining spirals of escalation.

    The problem posited by the essay, is that we (as a species) have gotten very very good at fighting the surrender acceptance instinct.

    There's the obvious case: With guns, it's possible to kill faster than the surrender reflex can kick in.

    ... but the essay also notes that a similar problem occurs when you start scaling conflict up. As the people with the agency to act take more strategic roles, removing them from the front lines, they no longer come face-to-face with those surrendering. The usual surrender acceptance instinct can not be triggered.

    While the essay is specific to conflict, I think there might be something here... not just about technology, but rather about scale. The coordination afforded at scale makes systems far more efficient. Otoh, it also removes those doing the coordination from the innumerable social cues that our species has spent eons developing around.

    In other words, I think it's deeper than technology rendering certain skills obsolete, or enabling cognitive offloading... The instinct to "be moral" is coupled in our species with an incredibly complex system of social signalling. This system is disrupted when we introduce technology as a mediator, and also when we start to operate at the massive scales technology enables.

    💬 1🔄 3⭐ 12
  • 💬 0🔄 0⭐ 1