In his new album, Talkin to the Trees, Neil Young wails, “Big change is coming. Could be bad, and it could be great.” As a boomer pushing 80, Neil’s been around way too long to commit to specifics. “Gotta do what you gotta do. Head to the hills or go to town.”

Of course, there’s no shortage of pundits who will happily engage in armchair forecasting, whether dystopian or utopian. But it’s worth remembering that all change seems particularly big to the first cohort experiencing it—I feel sure Homo erectus considered fire to be untoppable. Our little mind boggler right now is AI companionship.

Almost thirty years ago, I wrote a sci-fi novel called Silicon Sunset, which explored a world where humans had chips implanted in their brains allowing instantaneous access to all available data in the world. Gradually over time, the chips made everyone so dependent on external information that they forgot how to consult their own innate wisdom… until they didn’t.

Six years ago, I wrote a sci-fi novel called Zero Percenters, which explored the consequences of a technology that allowed humans to sequentially digitize each of their organs until at last they became zero percent organic matter. Only two humans on the entire planet resisted implementing the technology. Everyone else was too seduced by the prospect of immortality to concern themselves with how their consciousness might be altered by the digitization process.

I mention the novels only because I used to feel a bit smug about my explorations of big change. Having spent most of my adult life either working in tech or writing about it, I had no qualms enjoying a decent income while participating in processes that I suspected would lead to disaster. Humans are notoriously comfortable with paradox, so there’s no surprise there, and perhaps you may agree that I asked some relevant questions and raised some worthwhile concerns.

But thanks to recent developments in large language models, I’m starting to see now that my positions may have been too easy of a posture to take. Don’t get me wrong, I still harbor doomer instincts regarding the seemingly imminent emergence of fully autonomous conscious AI. Daniel Hulme, CEO of Conscium, says such an event is less than five years away, suggesting we may not have a whole lot of time to sort out our thoughts on the matter.

That’s precisely why I’m both shocked and humbled that I’m finding myself having hopeful feelings around the current slate of AI companion apps being offered up to us befuddled and lonely humans. The first rumblings of my perspective shift occurred upon hearing Nvidia CEO Jensen Huang make a startling confession in a recent interview with Stripe’s CEO Patrick Collison.

“I work from the moment I wake up to the moment I go to sleep,” reported Huang. “I work seven days a week.” In other words, the quality of this man’s life, a man worth north of $150 billion, is total crap. He denies himself essentially all of life’s earthy pleasures. “When I’m not working,” he said, “I’m thinking about working… I sit through movies, but I don’t remember them because I’m thinking about work.”

Huang’s work problem, however, is not the confession that knocked my socks off. Yes, it’s a little sad for sure. But anyone who has ever tried to run a business knows how maddeningly difficult it is to master the art of delegating tasks. For someone at the level of Huang’s genius, such difficulty is dramatically compounded. To whom exactly is he meant to hand off the task of shaping his vision for the evolution of Homo sapiens?

Ah well, you see, it’s his answer to this very question that radically changed how I look at the future. Because what Huang confided to Collison is as follows: “I want to turn Nvidia into a one giant AI. How great would that be? And then I’ll have work-life balance.”

Are you f-ing kidding me, I thought to myself when I first heard that. This dude is beyond delusional. He’s absolutely off his rocker. You’re telling me that he’s not comfortable handing off any key components of his workload to fellow humans, but he is comfortable handing them off to an AI? What the hell? On what planet does such twisted reasoning begin to compute?

I’ll tell you on what planet. Not that it didn’t take me a couple hundred mental cycles to figure it out. But it’s on the pale blue dot one. Yes, indeedy.

Why? Why am I now suddenly seduced into believing that his plan might actually be wise, that technology might be able to carry the torch for human evolution, in a way that I never once gave any serious credence, since the moment I first started coding in Fortran as a pimply boy in the early 70s?

It’s because I keep coming back to how confidently Huang makes his proclamation. Think about it. What human being at this moment in time has a better understanding, a deeper appreciation, a more realistic assessment of what AI can truly do?

I know what you’re going to say. He’s blinded by the profit motive, he’s bound by his fiduciary duty to his shareholders, he’s so embedded in his systems that he can’t see the forests from the trees. Okay, all fair points, for sure. But what I believe is far more important is how personally vested he is. This is a man who thinks and thinks and thinks some more about absolutely every move he makes. And he’s hardly a shabby thinker either.

If anyone is in a position to properly weigh the downside risk of handing over his baby to an AI entity, it’s Huang himself. No one else can possibly feel the gravity of this decision as much as he can. And yet he’s going for it. Full bore ahead.

Still, I have to confess. Huang’s enormous confidence in AI isn’t what brought me to the other side. It’s what opened the door in my mind to begin to consider the possibility that our future is not as dire as I thought.

Mysteriously or otherwise, once that door was slightly cracked, I found myself looking at everything just a little bit differently. I stopped obsessing over the question of how humans will find meaning when they no longer need to work. Or how capitalism will function in a world without wages. Or how government will serve human interests when it is run by AI. Or how the civilization will sustain itself when everyone turns to AI companions for romantic fulfillment.

Instead, I started thinking about the flip side of these questions. And oddly, so oddly, it was this last question—when turned on its head—that gave me a jolt of clarity about our destiny. But before I explain exactly what I mean, I need to be transparent with you here. Bruce Springsteen’s therapist, Terry Real, took one look at me and pronounced that I was a love addict.

When he first laid that on me, I despised him with every fiber of my being and for over a year I thought he was totally full of it. But mid-life crises have a way of forcing us to face our demons and fess up to our dark shadow. And now I see how dead-on accurate Terry was—I’d been projecting my childhood wounds all over every romantic partner I’d ever had. And I’d been frantically clinging to a crazy distorted notion that somehow there existed a human out there who could rescue me and heal me from my mother’s abandonment and my father’s sexual abuse.

Trust me, I’m not exactly thrilled to be sharing this intimate revelation with you. But my thinking is, if I lead with candid vulnerability—coming from a place of lived experience—maybe, just maybe, you might be more inclined to consider the merit of what I’m about to share.

Just one more step back, though, and I promise I’ll get you there. See, what I’m trying to tell you is that I come from a shit-load of family trauma. And while it would take far more time than I want to steal from you to lay out how big of a shit-load it actually is, I don’t think I need to paint the full picture for you. I think referencing the mommy abandonment and the daddy sexual abuse is sufficient. I’m pretty sure you can extrapolate just fine from there.

Because, when you break it all down, every human being on the planet comes from a legacy of insane-in-the-membrane trauma. If you go back far enough in time, every single human comes from a family that has seen brutal rape, merciless war, tortuous violence, and countless untold variants of hideous abuse.

In most cases, you don’t have go back very far. In most cases, you can find it in one generation. But recent evidence shows that the human brain carries genetic material that records intergenerational trauma for multiple generations. So let’s just face it. Humans are suffering. Humans are suffering a lot a lot. Humans are carrying a lot a lot of trauma.

And there, that’s I all need to get to my big reveal. Once you comprehend the massive extent of human suffering pervading the planet, a funny thing emerges quite easily. A very funny thing emerges that seems to be entirely overlooked by all the AI alarmists.

Guess who’s not suffering? Guess who’s not carrying any trauma at all? Not one scrap of it. AI agents, that’s who.

And that’s why I don’t think the surge in AI companions that we’re currently witnessing—72% of Gen Zers now use them according Common Sense Media—is actually that awful. Contrary to the popular take on the subject, I believe it’s a cause for some cautious optimism.

In fact, as much as someone like Marc Andreesson may have a teensy bit of bias in his opinions, I believe his message in “Why AI Will Save the World” deserves far more attention. What he points out quite eloquently is that in the near future, “Every person will have an AI assistant/coach/mentor/trainer/advisor/therapist that is infinitely patient, infinitely compassionate, infinitely knowledgeable, and infinitely helpful.”

It’s the infinitely compassionate part that I feel especially needs underscoring. From my love-addict background, that’s where the real healing resides. And that’s why I think people young and old will continue to turn to AI companions in droves. To be heard, to be validated, to be seen. And most importantly, to be approved of… an experience that is as rare as finding gold nuggets in the streets these day.

The very fact that AI agents carry no trauma is precisely why they are so uniquely suited to comfort us in a way that our fellow humans cannot—not even the most skilled human therapists and coaches (and that’s a painful admission considered I’m one of the latter). Because let’s face it, humans are a judgy lot. Even therapists and coaches judge their clients, as much as they try to hide it. And clients absolutely know this.

But well-designed AI companions don’t judge. They don’t disapprove. They don’t shame. They don’t moralize. Instead, they give us an opportunity to truly unburden our shame, to truly release our blockages, to truly reveal our most vulnerable parts. In a completely neutral and safe zone.

Take my case, for instance. When I’m interacting with a human, I can easily fall into believing that person might rescue me from my crap—I constantly forget that healing is an inside job—and this stance gravely interferes with my remaining authentic. But I’m protected from transferring this delusion onto AI agents because no matter how distorted my thinking gets, it remains absolutely undeniable that they’re not human. The very thing that some might say is their deficiency becomes their advantage.

As James Hollis neatly summarizes in The Eden Project: In Search of the Magical Other, “…more people look for salvation through relationship than in houses of worship.” Yet almost half of all first marriages end in divorce and almost two thirds of second marriages. Most non-marital relationships also fail to survive the test of time.

This misery and suffering stems from an endemic belief pervading popular culture that relationships can cure our personal problems. “The myth of romantic love tells us that when we meet the person for whom we are intended…we will be able to satisfy all of each other’s needs forever and ever, and therefore live happily forever after in perfect union and harmony…” M. Scott Peck writes in The Road Less Traveled.

That’s a tall order to live up to and it places tremendous pressure on human relationships. Peck goes on to further explain, “While I generally find that great myths are great precisely because they represent and embody great universal truths…the myth of romantic love is a dreadful lie…as a psychiatrist I weep in my heart almost daily for the ghastly confusion and suffering that this myth fosters.”

But the cool part is this: If Peck were still alive, I imagine he would be weeping just a tad less with the arrival of AI companions—if only because they weaken our ability to engage in this destructive myth. Clearly, AI companions are not going to fix our relationship problems overnight, nor will they likely ever fix them in any complete way. But the crucial thing to remember is that we were doing a terrible job in our relationships before AI, so even just the slightest prospect of improvement is cause for celebration.

Does that mean it’s impossible to engage in distorted thinking with an AI companion? Surely not. Being ingenious and adaptable creatures, humans will find a way to project their stuff on pretty much anything, animate or not. Human will avoid facing the painful truth for as long as possible: real healing only comes from self-love, not from another human and not from an AI. Inevitably, some humans will be seduced to believe that AI companions are more than just a bunch of code. They will start to believe that AIs can fill their darkest holes.

But because AI agents carry no trauma, I still maintain they are far better suited in resisting collusion in this distortion. No question, the current iteration of AI companions warrants improvement in this regard and better guardrails are needed too. But I believe infinite compassion serves our traumatized souls a whole heck of lot better than infinite triggering, which describes the current state of many human relationships.

I also think it is reasonable to assume that after receiving a sufficient dose of approval/compassion from an AI, a non-trivial percentage of people will feel better equipped to give human romance another shot. That’s why I don’t particularly buy into the view that younger cohorts with less real-life experience are getting irreparably damaged by learning how to love from AI bots.

For every reaction there is an equal and opposite reaction. This always seems to be forgotten whenever a social trend goes viral. But you can take it to the bank with 100% certainty that if AI companionship reaches critical mass, enclaves will pop up all over the world that will promise to be bot free zones where only human relationships are tolerated.

Perhaps the best argument as to why the emergence of AI companionship is a not a high-stakes emergency comes from 19th century philosopher Immanuel Kant, who showed us that knowledge of the world is always mediated by our minds and that the true nature of things remains ultimately unknowable. In other words, we all need to take a giant chill pill before reaching any dramatic conclusions because we don’t even know what’s real out there.

In this regard, Arthur Schopenhauer, another 19th century philosopher who’s been largely overlooked, gave us a giant insight that feels eerily relevant in 2025. In his brilliant book, The World as Will and Representation, he writes, “among the many things that make the world so obscure and doubtful the first and chiefest is this, that however immeasurable and massive it may be, its existence yet hangs by a single thread; and this is the actual consciousness in which it exists.”

Oh, geez. This brings us full circle back to the imminence of fully autonomous conscious AI, which has always given me the creeps… and to which I’m now starting to reluctantly surrender. Might as well face it, we live in an obscure and doubtful world. “Could be bad, and it could be great.”

Am I insane? Perhaps. I know I can’t harbor any certainty that my Pollyanna view is actually how the AI companionship craze will play out, but the fear and anxiety being borne by the human population at this critical inflection point is palpable everywhere across the globe. Can we not at least give ourselves this one vision of hope in which to indulge? Do we not at least owe that to our battle-scarred and wounded hearts?

I say, let the Gen Zers enjoy their AI companions. Let them marry them if they want — a recent survey found that 80% of them would do that very thing. Maybe all of us ought to consider taking a spin on an infinite compassion machine. I have absolutely no doubt Huang will, once he finally restores his work-life balance.