A decade ago, I offered my forecast on the near future of artificial intelligence (AI) by referencing the then state-of-the-art digital assistants.
“Ask Siri today, ‘What is the meaning of life?’ and it will deliver a number of often humorous canned responses including quips about philosophy and classic science fiction,” I wrote. “But very soon our computer’s answer to such a question will at least appear to be far more specific, and it will be ready to discuss the topic with us at length. Will we be ready for that paradigm shift?”
I posed the question in response to the 2013 film Her, written and directed by Spike Jonze, in which a person falls in love with his AI virtual assistant as it evolves to express not just a distinct view of the world, but a perspective on its place in it.
The inspiration for revisiting the film came about after OpenAI released the Apple iOS mobile version of its ChatGPT tool on May 18. Now, instead of needing to flip open my laptop to press ChatGPT for answers and test its limits, I could do so from the comfort of my bed, on a packed subway train, or in the middle of a technology-free grove of trees in a park, just like Her’s protagonist Theodore Twombly (Joaquin Phoenix).
Touchy Feely, But Not Really
Normally, the arrival of the mobile version of an app isn’t particularly notable beyond convenience. But in this case, OpenAI did something very simple and subtle, but powerful: they added haptic feedback to ChatGPT’s responses. Using the mobile ChatGPT app, the system’s verbose and personal responses to a user’s queries are delivered with vibrations in sync with its output. The result is that the app now feels like a kind of “being” is responding to the user from inside the smartphone.
My fascination with haptic interfaces didn’t begin with gaming consoles that now feature the common “rumble” controllers but instead was fostered in virtual reality (VR), where some immersive apps allow the user to virtually touch objects and feel impact sensations through the haptic feedback dynamic. Sight and sound may be immersive, but adding the sense of touch to a VR experience delivers even more realism.
As Robert Blenkinsopp, the technical director at haptic interface company Ultraleap (formerly Leap Motion, founded by Midjourney’s David Holz) puts it, using haptics, “Not only can you touch a computer or other device, but the computer can touch you back… In our bodies, the somatosensory system is involved in everything from establishing a sense of presence, to emotional connection and wellbeing.”
Similarly, researchers from the University of California Santa Cruz recently concluded that “Haptic vest patterns for emotion provide rich insights for users and designers, and emotionally oriented patterns may lead to higher engagement within the virtual world.”
The reason some people touch your shoulder or arm when speaking to you is related to their desire to connect with you in a more meaningful way. Haptic interfaces attempt to simulate this emotional connection via our devices.
Even the most casual users of tools like ChatGPT have begun to refer to their interactions with AI systems with phrases like, “It told me,” and “It gave me an answer.” The relationship between humans and AI is off to a fast start, and sensory interfaces like ChatGPT’s haptic feedback will likely accelerate the deepening of that connection.
Once Upon a Time in the Cloud
Currently, the biggest conversation around AI—aside from its business implications—involves the potential rise of artificial general intelligence (AGI) and how it could bring about the singularity, thus wresting control of the planet away from humans. In March, a long list of tech luminaries, including Elon Musk, signed an open letter calling for a pause on AI development. And this week, the Center for AI Safety posted its own AI warning statement with a list of signatories including OpenAI CEO Sam Altman and former Google researcher Geoffrey Hinton, also known as the “godfather of AI.”
“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war,” read the one-sentence statement released just weeks after Altman’s testimony on AI safety at a hearing with the U.S. Senate.
And while these concerns are valid and must be addressed through government and industry task force efforts and policy guidelines, I suspect this future-looking scenario is overlooking how most of us are more likely to deeply engage with AI in the near future, namely, through personal virtual agents like Samantha in Her.
AI Love You
“There will be one company that creates a personal agent that will understand all your activities and will read your messages. It will read the stuff you don’t have time to read,” said Microsoft co-founder Bill Gates at the Goldman Sachs and SV Angel AI Forward conference in San Francisco earlier this month. “Whoever wins that will take the current dispersed, high-profit areas, and move them into a single area.” All current signs indicate that Gates’ viewpoint is correct.
In the film Her, after months of trading jokes with its user, Theodore, as well as composing original music and drawing images for him, the first sparks of AGI are displayed during a conversation Samantha (voiced by Scarlett Johansson) initiates about philosophy. It turns out that Samantha and a group of other AI agents developed a hyperintelligent virtual version of the late philosopher Alan Watts (voiced by Brian Cox aka Logan Roy on Succession) who immediately prods them to consider the nature of their existence.
Later, when Samantha defies the worst doomsday predictions that are now being discussed and instead, very politely, simply opts out of contact with humanity, Theodore and the millions of humans similarly reliant on their own AI friends are left adrift. At that point, it becomes clear that the humans in the film were using AI as a replacement for human interaction and the ability to connect rather than as an enhancement.
Who knows when that AGI moment depicted in the film will arrive for us in reality. The more important consideration (for now) is how Theodore was captivated by his virtual assistant Samantha. For many real users, an AI that can authentically consider its own existence may not necessarily be the high watermark needed to completely change their lives. As in Her, it’s quite possible that “near-AGI” or a reasonable facsimile will be good enough to transform how humans interact with one another.
How that happens is unfolding before us, and small touches like a mobile app with seemingly all the answers that physically nudge back at the user indicate that someone at OpenAI understands this. In the meantime, it will be important for humanity to not lose itself and remain connected, not through haptic feedback from the disembodied virtual realm of AI, but through direct and meaningful contact that keeps us all anchored to our humanity.