A.I.: All Too Human, None Too Human

Musings from a Eulogite at London Tech Week

13th June 2019

By Mohammed Sheriff

For the longest time, the most transcendental flights of human fantasy found their place in fiction, superstition, myth and fables. Both fascinated and terrified by what is not quite human – either going beyond it or falling just short – we invented gods and demons, superheroes and zombies, paradises and infernos.

Existing only in the imagination, these stories have functioned as the engines of civilisations through the ages. They became reference points for the group identities, systems of government, ethical norms, art and entertainment that shaped history’s course.

Now, our unique talent for dreaming up eerie beings has found its corporeal form. We have breathed life into the lifeless. Simultaneously human and non-human, inspiring awe and revulsion in equal measure, it has a new home in the unfolding technological revolution.

Its creations are not restricted to belief systems agreed upon by a tribe, community or nation. It demands no leap of faith and requires no suspension of disbelief. Its presence is reaffirmed in almost every facet of our daily lives: the algorithms sitting under our social media platforms; the outputs of industries as far apart as transport and marketing; the instructions we casually issue Siri and Alexa to meet our most banal needs.

What’s remained constant is the visceral way in which we react to the birth of these spooky children of our unrelenting imagination. With each technological development, we either jerk toward utopian dreams of eternal abundance, or dystopian nightmares of subjugation to a greater force culminating in extinction.

And such claims – at both ends of the spectrum – are by no means confined to the idle speculation of crackpots, soothsayers and doomsday prophets. The philosopher Nick Bostrom hypothesises that a certain level of technological advancement will, by default, lead to devastation on a civilisational scale. By contrast, Ray Kurzweil, of Singularity fame, predicts that medical technology will one day deliver the fountain of immortality from which we will all drink.

Whether or not any of these hypotheses, predictions and thought experiments are credible, they are unlikely to ever be dispelled. This is especially true when they are formulated by some of the most highly regarded minds of our time.

For tech companies at the frontier of innovation, this presents the uniquely challenging task of managing expectations of potential rewards and pacifying concerns of potential catastrophe. The commercial imperative to attract investors and appeal to customers has to be balanced with a clear message that this is all to humanity’s net benefit – rather than cost.

After all, much of the worry is that either humanity will lose out, or that we will lose our humanity. This is the arresting insight an exhibit at the Barbican Centre, AI: More Human Than Human, intends to provoke. We want reassurances that we will always remain just as human and are disconcerted by the prospect of creating anything more human – and by extension, of more significance than ourselves.

At London Tech Week this year, it became clear that tech firms know this is a core issue they will have to address. In a session titled ‘Making Computers More Human’ at CogX 2019, presenters from companies including Emotech, Soul Machines, and Constellation AI each emphasised that their main goal was to enhance the human experience rather than replace it.

It’s a compelling message, and it’s hard to doubt the sincerity of those who issue it. To a sympathetic crowd of tech enthusiasts, it largely has the desired impact. The next step is to bring the rest of humanity into the fold – and then deliver on the promise.