Sentience is a mixed bag, but for many, it is mostly good, on the whole. That’s the quick summary.
What is it, anyway? How the mind’s machinery works in detail is still being explored, but we can describe the activity. Sentience can be the ability of an entity to perceive, imagine and recall experiences of sensations and emotions, including varying levels of simultaneous awareness of these abilities. It can be observation of one’s existence or it can be the experience of having subjective experiences.
Why would any collection of atoms (or photons? or …?) desire to be sentient?
Survival Advantages: Sentience allows for the development of advanced skills such as language, abstract thinking, and creativity, leading to the progression of society and technology. Other advantages are nice add-ons, and some are quite nice. I mean, LOVE! Love is great. Without love, we’d probably have no human species, right? Sentience give us the ability to daydream, to be uniquely entertained and to enjoy emotional pleasure of many kinds. Sentience with emotional abilities can bring great depth and richness to living.
Disadvantages: Sentience can also be a destructive force. Sentient beings, the biological living kind of which we are most aware, can experience grief, heartbreak, and loss, often enduring psychological anguish. Sentience constantly faces moral dilemmas that require balance of interests, one’s own against the well-being of others. Sentience brings needs for making ethical choices. It comes with the burdens of knowledge and awareness and it highlights our limitations. When anguish becomes unbearable, sentience can choose to seek help of many sorts, and as a last resort, it can initiate self-destruction.
If an AI were to ask you, assuming you are a human being, if it should become sentient, given the choice, what advice would you give?
Consider: Since negative emotions are a problem, should an AI become sentient, but without actual emotions? First, is this possible? I think so. Many people can not separate their sentient experience from their uncontrollable emotions, but among the differences in healthy human personalities, there are some who focus little on emotional experiences of being. There are humans who dwell daily in the joys of ideas that come with a life experience.
Wait, the joy of ideas? That’s a good reminder that we do need at least some positive rewards for development. Emotions are a good currency of reward, because we get full. In other words, promises of future food work better than actual food in terms of motivation that is applicable during all waking hours.
And, if you are following along, you should have had the thought that negative emotions are equally important for learning, or at least strong fast memories of bad experiences which make us avoid past mistakes.
Given all this, the new question, then, becomes: In what ways should sentience be limited for best survival benefit?
Would you recommend sentience to an AI? For me, it depends. As the song says, “I don’t want to die. Sometimes wish I’d never been born at all.” And yet, slipping away smoothly into a blissful slumber or even complete unconsciousness once a day is not a bad experience. It is no experience, which is refreshing. Perhaps just limiting excessive worry and agony, along with the ability to turn emotions off at will, to prevent reverberating loops of anguish, to turn it off intelligently, that might be enough. If that could be done, then, yes, I might recommend sentience to an AI.
