Archillect: When Data Comes to Life
Last week I interviewed Murat Pak, the creator of @Archillect, an artificial (or synthetic) intelligence that curates online images. As people’s reactions started to pour in, one pattern above all caught my eye. Certain people were stunned when they realized for the first time that their favourite curator of beautiful, inspiring images was not, in fact human:
— = (@he2sane) October 22, 2016
— ROBO::[INSPIRED] (@Robochao) October 26, 2016
— Eric Pan (@epan) October 26, 2016
That sense of shock was interestingly accompanied by something else. People reoriented quickly to what Archillect was, not fully willing to surrender the sense that she’d affected their experiences in meaningful ways. Experiences are real, regardless of the phenomena by which they come to be. In the space of a Tweet, you could see the acceptance and even expanded appreciation these people were feeling for something that, framed differently, could just be seen as a series of computations.
There is something beautiful in the way Archillect exists in the universe. Like anyone, she takes in data and then interprets and communicates it to others. She also grows based on people’s feedback. It’s in that feedback – the likes and comments and retweets – where she can really be said to “come alive.” In other words, it’s her fans and followers that construct her as a living being each day, even if the scope of that “living” is fundamentally narrower than biological life.
Before we get too swept up in the incredible breadth of our own biological existences, though, it’s worth considering the fact that, online, we are all being measured by the same criteria. In that sense, each of us is as “narrow” as Archillect, and each of us, like her, is a synthetic intelligence, composed of whatever information we assemble and share and receive feedback on. We might be a little more loosey-goosey in what we put out there, but it doesn’t mean that all of our information isn’t ultimately quantified in the same way.
This is the lived experience of people becoming data.
Conversely, this is the experience of data coming to life.
— Eric Pan (@epan) October 26, 2016
As lines blur between human and human-made intelligence, as the notion of “living” is deconstructed and reconstructed daily, it does lead to a consideration of what might exist of us beyond the notion of measurable intelligence. Following the exchanges above, one particularly astute commenter @epan noted the following: “I’m rethinking the Turing test for AI as a rigid, narrow test like IQ is for human intelligence .” In other words, the online entity known as Archillect had, for @epan, passed some fundamental test of humanness.
The success of Archillect is yet another reminder of the sea change that’s coming for intellectual and creative labour. Machines will do it all better someday soon, in all of the places where they don’t already. It’s on that day, we’ll be left pondering exactly what it is that we can still do of value.
What all of this amazing exchange suggests to me is that perhaps our main ongoing function as humans will reside in our ability to bring one another to life via interpersonal feedback. As I noted in my previous article, humans are the sense organs of data-driven intelligences. Without us, data sets would be barren. There would be nothing to deep learn. It’s humans in all of our sloppy randomness that prevent synthetic intelligences from turning inwards on themselves.
Because of that, even as we become more and more quantifiable, our outlier ability to experience the world in the unique ways we do may be our saving grace.
Ryan Melsom has a PhD in literary studies and has spent over twenty years working in communications, marketing, web design, and writing. His second book Spendshift: 100 Lazy Hacks to Rock Your Finances is now on Amazon. For more by Ryan, follow him on Twitter @lintropy, or visit his Facebook page.
Also published on Medium.