THE COMPLETE LONG-FORM ESSAY
What If Generative AI Is Only the Beginning?
From human imagination to creating life beyond Earth
What if the first civilization we establish on another planet doesn't have a single human in it?
That question began as a thought about artificial intelligence. It led me toward robotics, biology, autonomous spacecraft—and eventually back to something much closer to home: the human need to keep creating.
We build businesses, develop technologies, solve problems and reach milestones. Then we look for the next challenge. I have been wondering whether the things we are learning to create could someday become part of a journey far larger than our own species.
This essay brings together my seven-part “What If?” series. It starts with capabilities that exist today, then follows a chain of increasingly speculative possibilities. The final question is philosophical: what if our ambition was taking us somewhere we do not consciously understand?
When intelligence learns to create
Generative AI has made creation feel unusually accessible. A thought can become an image, a piece of code, a design or the beginning of an explanation. The distance between imagining something and producing a first version has shortened.
What interests me most is what might happen after that first version.
An image of an unfamiliar creature is an act of imagination. A creature capable of living is a physical system. It needs a workable chemistry, a source of energy, a way to maintain itself and conditions in which it can survive. A convincing picture tells us very little about whether any of those requirements could be met.
The same distinction applies to a machine, a material or a medicine. Generating a possibility begins a process of testing it against the world. Nature does not become more forgiving because a design looks persuasive on a screen.
That is why I see generative AI as an opening chapter. Its larger consequences may emerge as it connects with tools that can measure, experiment, manufacture and learn from physical results.
Giving intelligence a way to act
Robotics begins to close the gap between a digital idea and a physical action. Systems that combine perception, language and robot control are already being developed. Google DeepMind’s Gemini Robotics, for example, connects AI models with actions performed by robots.
These systems are not universal machines that can reliably do anything. Their significance is more specific: intelligence is increasingly being connected to bodies, instruments and tools.
Now add manufacturing. A system that can inspect a component, identify a problem, produce a replacement and test the result has a different kind of reach from a system that can only describe a solution.
Even on Earth, making those steps work reliably together is difficult. Machines need energy, materials, maintenance and a supply chain. A truly independent system would have to deal with failures that its designers had not anticipated.
But the direction is worth thinking about. If intelligence can help decide what to build, robotics can manipulate the world, and manufacturing can produce useful objects, how far could that combination eventually go?
When the design problem becomes biological
Biology adds a much more demanding dimension to that question.
Researchers already use CRISPR to modify DNA in living organisms. AI systems such as AlphaFold can predict protein structures and help researchers investigate molecular interactions. These are substantial advances in our ability to study and work with life.
They do not amount to the ability to invent any organism we can imagine.
Editing existing biology is different from designing a complete living system. A prediction is different from an experimentally validated result. A genome is part of a living process that also depends on cellular machinery, development and its environment.
Still, I find the convergence fascinating. Imagine increasingly capable AI working alongside robotics, automated laboratories, synthetic biology and manufacturing. Could such systems someday do more than help us understand existing life? Could they eventually help create forms of life adapted to conditions very different from our own?
That is a much greater leap than the technologies we have today. It is a question about a possible future, rather than a capability we should assume is inevitable.
Sending the capability to create
We usually imagine interstellar settlement as a journey made by humans. We picture spacecraft carrying people, habitats, food and perhaps the familiar organisms on which we depend.
But humans are deeply shaped by Earth. Our bodies need a narrow range of conditions. Keeping people alive across interstellar distances would be an immense undertaking.
So consider a different kind of spacecraft.
It has no human crew and carries no living organisms. Instead, it carries AI, robotic tools, manufacturing systems and digital information. At a distant destination, it first studies the world it has reached.
What chemistry is available? What sources of energy could be used? What materials can be processed? Are conditions stable enough for any proposed system to function? And, before attempting to change anything, what might already exist there?
Digital instructions alone would not be enough. The craft would need functioning machinery, suitable feedstock, energy and the ability to carry out an extraordinary sequence of physical operations.
With technology far beyond our own, could such a system eventually attempt to create life suited to its destination?
Related ideas have appeared in theoretical proposals, including Claudius Gros’s Genesis Project. They remain proposals. The version that captures my imagination is a system carrying the possibility of a beginning, rather than a ready-made civilization.
Perhaps we do not need to send civilization across the stars. Perhaps we need to send the capability to create one.
Different worlds could mean different life
Now imagine many such craft, each travelling toward a different world.
Their destinations would have different atmospheres, temperatures, chemistry, radiation and sources of energy. A form of life suited to one environment might be unable to survive in another.
The interesting possibility is therefore not simply copying Earth everywhere. It is the idea that very different environments could support very different beginnings—if the underlying science and technology ever made that possible.
Even then, a living cell would not be an ecosystem. Survival would not guarantee complexity, and complexity would not guarantee intelligence.
Evolution has no obligation to produce a civilization. It has no established destination. Across enormous timescales, life can diversify, persist, change or disappear.
That uncertainty is part of what makes the thought experiment compelling. The eventual outcome might bear little resemblance to the species that set the process in motion. We might never see it. We might not even be able to recognize what it becomes.
The question that turns back toward Earth
Once I follow that possibility far enough, the direction of the question reverses.
If a future civilization could attempt something like this, could an earlier civilization have considered a similar idea? Could the creation of life be part of a much older sequence of beginnings?
The possibility is not evidence. We do not have evidence that an earlier civilization deliberately seeded life on Earth. A story that is imaginable does not become an explanation simply because it is intriguing.
For me, its value is as a philosophical lens. It challenges the assumption that humanity must be the final destination of the story we inhabit.
We can be important to our own experience and still be one stage in a much larger history. Thinking on that scale changes the questions I find myself asking about progress, creation and responsibility.
Why are we so relentlessly driven to create?
This is the part of the journey I keep returning to.
We finish a project and begin another. We reach a milestone and move the goal further ahead. We discover something, then immediately want to know what lies beyond it.
Much of that drive has familiar, immediate purposes. We want security, opportunity, recognition, understanding and better lives. Yet the feeling of “what next?” can remain even after we achieve what we once wanted.
I have been wondering whether our restless ambition could be taking us somewhere we do not consciously understand.
What if something had been seeded deep within our subconscious, pushing us to evolve, discover and create? What if the technologies we are developing were steps toward an outcome much larger than making our own lives more convenient?
That question does not turn human dissatisfaction into evidence of a hidden design. I mean it as a philosophical possibility, not a scientific conclusion.
But it offers an unusual way to look at our place in the future. Perhaps our story will be judged less by the milestones we accumulate than by the possibilities we make available to whatever comes after us.
Are we the destination of evolution, or merely one step in a much bigger journey?
What will we create?
I began this series thinking about machines that generate content. I ended up thinking about the possibility of creating life, and about the motives of the species building the machines.
There are enormous gaps between those ideas. Some may never be crossed. The point of following the chain is to see what new questions appear when developments that usually sit in separate conversations are considered together.
If our creations could someday reach places we cannot, adapt to conditions we cannot survive and begin stories we will never witness, then the meaning of human achievement would become much larger than our own experience.
That possibility leaves me with the question that closes the series:
What if our ambition was seeded with a purpose?
Maybe our greatest achievement won't be what we become. Maybe it will be what we create.
Further reading
Gemini Robotics brings AI into the physical world — Google DeepMind
AlphaFold — Google DeepMind
CRISPR — National Human Genome Research Institute
Developing ecosystems on transiently habitable planets: The Genesis project — Claudius Gros · theoretical proposal