In the shadow of generative AI, what remains uniquely human?

Generative AI is rapidly changing the way people work and live. Through language replication and the generation of written content, images and even music, gen AI is encroaching on domains previously considered ‘uniquely human.’ As the verbal and cognitive capabilities of machines evolve, an existential question has emerged: In gen AI’s shadow, what unique qualities will humans retain?  

More than 50 years ago, Stanley Kubrick’s groundbreaking film 2001: A Space Odyssey gave moviegoers and society at large one of the first glimpses into the future of AI. In the movie, a spacecraft’s onboard computer verbally interacts with its human crewmates, executes all technical aspects of the mission and even plays (and wins) a friendly game of chess with an astronaut. At one point in the story, the computer — HAL 9000 or simply “Hal” — is interviewed remotely by a news reporter back on Earth.

Moments later, when the interview shifts back to the crew, the reporter says he felt that Hal exhibited a sense of pride when he spoke about his own technical flawlessness. When the reporter asks if they think Hal is capable of experiencing emotions, the mission commander is doubtful.

“Well, he acts like he has genuine emotions… but as to whether or not he has real feelings is something that I don’t think anyone can truthfully answer.”

More than a half-century later, the ability to experience feelings and emotions are qualities that — at least, for now — remain uniquely human.

Interestingly, and contrary to Hal’s conversational abilities in a fictional setting, language was not something computers did particularly well. Today, however, gen AI has revolutionized natural language processing (NLP) tasks that include large language model (LLM)-driven language translation and sentiment analysis, and chatbots can now understand and respond to questions and commands. In a particularly noteworthy example, AI-enabled a computer to pass the Turing test while simultaneously convincing multiple human judges it was a person and not a machine.

Beyond the purview of technology

As gen AI continues automating human tasks without “feeling” any particular way about doing so, those of us among the living can take stock in our other unique qualities that are incapable of being mimicked by machines. Along with emotions, attributes that remain uniquely human include imagination-based creativity and original thinking, and complex problem solving that requires cognitive flexibility and intuition. It’s also important to note how morality and ethics — which are beyond the purview of a technology that lacks an experience of being a member of society — factor into human decision making.      

The five human senses, and the extensive input the brain processes in relation to them, represent another example of what can be deemed uniquely human. As sight, sound, smell, taste and touch intertwine with an infamously fallible memory to create an embodied experience in humans, it’s difficult to imagine technology replicating the uniquely human experience of this convergence of senses.

Delving deeper into the incomparable characteristics of the mind, the discovery of mirror neurons represents another human characteristic that technology has yet to reproduce. A mirror neuron is a neuron that fires both when an individual executes a specific motor act or experiences an emotion and when they observe the same or similar act performed or emotion experienced by another individual. First observed in primates, actions driven by mirror neurons can be described in the most simplistic terms as “monkey see, monkey do.”  

According to research published by the National Institute of Health on mirror neurons, “from a functional viewpoint, action execution and observation are closely related processes, and… the ability to interpret the actions of others requires the involvement of our own motor system.” These mirror neurons enhance our experience of empathy, competition and teamwork, to name just a few examples. While an LLM might infer what we are feeling, they do not feel it themselves.

A shift in organizations’ mindset

In tandem with gen AI’s rising trajectory and the thought-provoking existential questions that come with it, humans are grappling with how to manage, control, and regulate AI technologies. Going forward, organizations will need to make choices when delegating tasks to gen AI technology.

According to research by McKinsey, business leaders need to take a broad view of gen AI’s capabilities and “deeply consider its implications for the organization.” The findings revealed that many global executives shared the following sentiment: “We were behind on automation and digitization, and we finally closed the gap. We don’t want to be left behind again, but we aren’t sure how to think about generative AI.”

Seeking to not repeat missed opportunities of the past, many organizations are approaching gen AI cautiously. Companies that leverage gen AI will need to establish well-defined workforce implementation and utilization strategies to ensure the responsible execution of their adoption roadmap. This will be increasingly critical as new regulations are created to ensure gen AI is used ethically, and standards are established to ensure data privacy and security. In short, organizations with a legitimate stake in gen AI will be held accountable for how they develop and deploy it.

Just because technology can do something…

Traditionally, technology has exerted a heavy hand on what we consider “work,” as 60% of the job titles held by those employed in 2018 did not even exist in 1940. As we look ahead to a world increasingly mediated by AI, it remains to be seen what new endeavors humans will undertake as AI remakes the 9-to-5 landscape. Going forward, governments, corporations and organizations of all types will need to make critical, conscious decisions about what will be outsourced to computers and what roles will remain in the human realm. During this process it’s important to consider this: Just because technology can do something doesn’t necessarily mean it should do it. 

When futurist Arthur C. Clarke and Stanley Kubrick partnered in the 1960s to write a screenplay that placed AI at the center of its plot, could they have known how prescient their fiction would one day be?

Richard Sonnenblick is chief data scientist at Planview.