The dawn of A.I. companionship has been on the gradual rise for a very long time. So long, that I think it’s fair that nobody has figured out how to hit that “sweet spot” yet – the perfect balance of functionality and compatibility. How do we make it human, with none of the faults?
The funny thing is, the majority of people actually interested in A.I. companionship are not interested in this at all. Happy suburban families or ambitious up-and-comers without a cent to their name have no interest in the investment on A.I. companionship. They are quite satisfied with Alexas and Siris, who add a little bit of convenience to their daily lives but not much else. I personally use my Siri for little more than a lullaby machine playing distant train sounds, an alarm clock that can wake me to smooth jazz on iHeart Radio, and a bluetooth speaker. Thus ends my imagination with such a “companion” device.
But now we are on the cusp of true A.I. companionship, teased through the programs used to draft writing and art through the input of carefully coded keywords. The human trick in A.I. that makes it desirable is not its perfection, or its speed, even, but its adaptability. A computer, after all, will only put out what you put into it. But it will retain that knowledge, so that, the next time you ask for something, its response will take your previous action into account. A.I. does not work as “one-size-fits-all.” It must, as humans do, begin with a child’s scope, adapting to what it thinks will please us, the parent.
Consider the “Companion Robot” in the video above. Stilted and unexpressive, breaking silence and lack of interaction with the same cutesy hand wave motion several times in a row. It is not programmed to learn or adapt with its owner, but to fit a stereotype that many lonely individuals might find appealing. It is a harmless, unimpressive display (literally just a display, less of a computing mind) when compared to what could be. Like the RPG choose-your-route dating apps; all of it is scripted, a farce you act out under the false pretense that the program is reacting to you. A.I.’s new potential in 2024 has the possibility of mimicking human neuroplasticity, proving that perfection is built gradually, not shipped out at conception. That is not to say that this wouldn’t require several paid updates to achieve, however.
So, what’s stopping us? The biggest issue, besides the mires of cybersecurity and legal precedent, would be target audience. I know it came off that I was certain what people were looking for, but one of the most deceiving characteristics of a human’s relationship with their own neural pathways is what they truly want. Do they want what is easy, predictable, restrained? Or do they want what is complex, spontaneous, raw? That we still have such a hard time mapping these desires out in relationships with other people means that companies and the A.I. themselves will find themselves at odds with what they think the customer wants. Worse yet, there is the possibility that this contentment in fruitless endeavors and relationships with an A.I. companion will wear off, or perhaps not even take on. For the psychological stimuli of “happiness” is often accompanied with a lasting physical reward registered by the one of our five senses. No hologram can stimulate more than two senses, and the brain will never accept such a pseudo-existence as a complete companion.
Morning's Early Dark 