If Meta CEO Mark Zuckerberg is any measure of how realistically we are taking artificial intelligence’s looming threats to humanity, the answer is: not realistically enough.
For the past few years, AI has hurtled towards each next step in optimization. ChatGPT made for a mainstream tool, but before that, people speculated — as in movies such as Her and iRobot — about AI sentience. We haven’t gotten there, and in actuality, we won’t. But concerns over the unknown tend to preoccupy society, especially over technology coming from self-professed utilitarians. Right now, these concerns center around job replacement and child harm.
Human disconnect is the broadest of our AI fears — probably broad enough to encompass most others. It has the curious quality of sounding more banal and more forgettable than imminent unemployment crises, as well as more distant than the obvious problem of teenagers’ social media addictions. Yet it is also the most terrifying, probably because it is the easiest to write off as impossible. So when questions about AI chatbots replacing human friends come up, it seems that most of us are too far gone to see that it’s already happened.
That group includes and perhaps is led by Zuckerberg. In an interview with podcaster Dwarkesh Patel, the tech executive noted some of his plans for Meta’s AI chatbot capabilities. They have already deployed, and as such have already seen scrutiny over their tendencies toward the sexually explicit. Zuckerberg, however, views the chatting tool in terms of meeting people’s “demand for” more friends than they have.
It is somewhat of a necessity that Zuckerberg does not see the AI friendship tool as something to replace human friendship, and so his “default is that the answer to that is ‘probably no’.” In fact, Zuckerberg sees only the upside: “As the personalization loop kicks in, and the AI just starts to get to know you better and better, I think that will just be really compelling.”
The reality, of course, is that technology has been of serious detriment to simple human relationships for years. One could look before the smartphone, but that is the clearest inflection point. Attention — both span and gaze — were absorbed into the screens, and social media emerged immediately after as a nail in the coffin. Generation Z users lament all sorts of unknown nostalgia. People spend less time together, and even get less out of the time they do spend together. To argue that the much more captivating and complicated tool of the AI chatbot won’t cause these conditions to worsen is not daftness, or even denial, but malintent.
EUROPEAN UNION WATCHDOGS FINE META, APPLE FOR ANTITRUST VIOLATIONS
It doesn’t take much to realize that we might want to hold onto the threads of pure human connection we have left, and for someone so entrenched as Zuckerberg, his comments speak mostly to callousness: Hearing after hearing on the harms of his social media platforms have not inspired any lasting convictions.
The technical, study-backed mental health outcomes of this specific technology have yet to be published. But it’s intuitive: What we have already seen, from tools with much less to offer, is terrible. That is harder and harder to recognize the more we forfeit our intuition to AI. But behind the machine learning, there will always be the simple choice not to endow chatbots with such human dead-ends.