Virtual Life

While waiting for my daughter in San Francisco, I was reflecting on the progress of technology—especially the achievements and future of Artificial Intelligence (AI). AI will likely receive a powerful boost once it's installed in robots. Robotic AI will learn about the physical world and improve by making mistakes as it performs various tasks—running factories, working the land and growing food, managing forests and cultivating timber production, assisting with scientific research and publication, and working in the service sector. Robots will also build new generations of robots, pass on their experience to them, train them, and build new factories. The only limiting factor is energy. It constrains the accelerating growth of the entire AI-robot ecosystem. And that might be a good thing, as the production of material goods continues to grow. Eventually, I thought, human labor will be limited to creative work and spheres where human-to-human relationships are essential. Likely, only volunteers will remain in the workforce. But I also worried: could an overabundance of material goods and services halt human progress? Might it rob people of ambition, making them lazy slackers? The apartment-hunting for my daughter in San Francisco unexpectedly enlightened me—no, AI can't fulfill all human desires. Even materially let alone spiritually.

I was walking with my daughter in San Francisco’s Marina District, where we plan to buy an apartment. The two-story homes were built in the last millennium but are beautifully renovated. A sunny coastline with yachts, the Golden Gate Bridge to the left, Alcatraz Island in the distance to the right—it’s a dream location. No new homes have been built here for a long time, and expansion is prohibited. AI cannot multiply housing in desirable living locations. One of life’s core aspirations—living where you want—remains unsolved. Land is finite, and most people don’t want to live in high-rises. In many beautiful areas, they’re banned anyway. A house in the Riviera, San Francisco, London, or Paris can only be bought from an existing owner. But who will sell? Who needs money in a post-scarcity economy? Real estate has become a status symbol—eventually, it will be unaffordable. We had to act quickly if we wanted to buy a condo for my daughter.

This led me to a new idea: could people be satisfied by simulating desirable places through Virtual Reality (VR)? I shared it with Greg R., a director at Google’s VR division.

We met in a Palo Alto Starbucks. I asked how VR was progressing.

“VR and AR are no longer trendy—AI has taken over. But companies keep improving it, especially for gaming. With AI, we now simulate not just visuals but sounds, smells, and temperature. Virtual touch is next—you’ll be able to feel and move virtual objects,” Greg said.

“But will people be able to eat virtually and feel full? Swim and feel the water? Interact with others? Sleep and wake up in the same virtual place?” I asked.

“Not yet. That’s still science fiction,” he laughed.

I explained: “Real estate in coveted locations is finite. Demand will keep growing. VR could let people live virtually anywhere—any city, district, home. Even if energy-intensive, it would be far cheaper than real real estate. That’s a massive market—nearly the entire world.”

That’s how it started. Greg and I founded Desired Living LLC. Google invested in equity. Andreessen Horowitz followed. Our team grew fast. Our VR products were powered by AI trained like self-driving cars and factory robots. In parallel, we have funded memory-scanning research at UCLA’s Brain Research Institute. The goal: recreate any stage of a person’s life virtually. This became the Virtual Youth (VY) project. By aggregating memories, we could recreate an accurate model of the past. Eventually, the entire living memory of humanity could be restored, returning time to a past that was still remembered. One could virtually live in that past.

Our first products, Super-Reality and Forever-Young, launched just before our IPO. Sales soared. So did our stock. But within a year, Super-Reality users grew disillusioned. Everything seemed real—sights, sounds, smells, climate—but it was still a copy. Forever-Young faced the same issue. The past didn’t feel authentic. People deeply value originals: art, heirlooms, famous racing cars, first editions etc. They preferred real San Francisco or London to perfect digital copies. Dissatisfied users posed a real challenge. We asked ourselves: what separates originals from copies? Often, only knowledge—knowing which is which—and the emotional weight that knowledge carries. The problem wasn’t technical. It was emotional. We had to replicate the emotional pull of the originals.

We consulted UCLA’s neuroscientists. Could emotional memory be decoded and recreated? Informational memory was easy—images and sounds are mostly objective. But emotional memory is subjective, shaped by personality. Studying it required a team of neurologists, psychologists, and philosophers. We funded the research. No one expected fast progress. Meanwhile, scientists began scanning and archiving volunteer memories, even those they couldn’t yet interpret. The idea: when analysis methods emerged, emotional memory could be extracted. Time was of the essence, as memory is irretrievably lost when people die.

A breakthrough came a decade later. Emotional memory was decoded. Surprisingly, it extended even before birth. Children remember stories told by parents and grandparents—emotionally charged and formative. These memories shape them. Reconstructing emotional memory allowed us to go back beyond one generation.

But isn’t all memory emotional? We see, hear, smell, and touch through our senses, but the result is always an emotional reaction. Emotional memory turned out to be much broader than informational memory. By restoring emotional memory, we restore the world that evoked those emotions. Subjectivity was a challenge—we all remember the same events differently because they provoked different emotional responses. But there were many shared patterns. The differences added nuance—like seasoning on bland food.

A global project to scan emotional memory began. Most people volunteered, hoping their memories would preserve their identity forever. AI learned to read emotional history by measuring brain and body responses to stimuli. Humanity’s memory archive exploded.

With emotional memory mapped, we vastly improved Super-Reality and Forever-Young. It wasn’t enough to simulate environments—we had to simulate how people emotionally responded to them. But it wasn’t easy. Emotional reactions varied wildly. Some stimuli meant to trigger joy caused distress. A trip to the past wasn’t guaranteed to be pleasant. Nor was a virtual dream home always satisfying. But there was no mass disappointment—no nostalgia for the original. We improved our products based on these negative experiences and the products were a hit.

I realized that progress never stops, even in an era of abundance. New research in our company’s labs aimed to improve the past or virtual reality using AI. Specialized realities began to emerge, tailored to individual needs. Custom worlds. Our company has grown exponentially. But my enthusiasm waned. I was tired of progress. I kept working out of inertia. Then I decided to step down from active leadership. I watched from the sidelines as almost all of humanity gradually moved into virtual reality. Each person, their own reality.

My daughter still lived in a real home in San Francisco’s Marina District. But at times, I felt she lived in virtual reality. I brushed the thought aside, but it kept returning. And what about me? Do I live in the physical world? More and more, the thought haunted me that we not only create the world in our imagination but may ourselves be the products of someone else’s imagination. The world is a subjective reality that doesn’t exist in space and time—but only in our imagination.