In my previous articles on “Should we still want biological space colonists?” (part 1, part 2) I’ve considered the question:
Should we still want to send human astronauts to colonize space? Or should we want to leave space expansion to AI?
I'll refer to this question as "the question."
I'm persuaded that human-level and then superintelligent conscious artificial intelligence (AI) will take over in the long run. They will spread into the universe. We’ll merge with them to some degree. But the perception of a difference between ‘us’ and ‘them’ will gradually fade away: all will be ‘us’.
I've seen angry reactions from many people who don't like the idea that our AI mind children will take over and bring the flame of consciousness out there among the stars. They insist that we, organic humans 1.0, must continue to lead.
Cosmists and Terrans
Reflecting on how to call these two opposite philosophical convictions, I think the best names are still "Cosmists" and "Terrans." These names were introduced by Hugo de Garis in "The Artilect War" (2005). "Artilect" is de Garis' term for ultra-AI billions or zillions of time smarter and more powerful than us.
In "The Consciousness Explosion" (2024), Ben Goertzel explains that Terrans are "people who place humanity at the fore and consider the good of humanity as the ultimate value."
Cosmists, on the other hand, are "people who are more interested in the flourishing of intelligence and life and complexity throughout the universe than in humanity as a particular species."
Ben calls himself a Cosmist, in a wider sense that includes this, and so do I.
Hugo de Garis foresees a big conflict between Terrans and Cosmists, culminating in a big war that will kill billions.
I hope we won't come to these extremes, but it seems evident that the conflict will be big.
Just think of the transgender issue around which we see all that drama and hate speech and political bickering. What will we see when it comes to the threat of a conflict between humans and hugely superior artificial beings?
One often hears the point that there's a third way: human-AI hybrids, or cyborgs. Future humans could have the option to merge with AIs and become cyborgs. Then perhaps the conflict between Terrans and Cosmists won't be that big?
"Hugo saw clearly that cyborgs were not a resolution to the Terran/Cosmist dichotomy," notes Ben. AIs can be a huge number of times more intelligent than humans. Then "what sense does it make to hybridize an AGI brain with a human brain?," Ben asks. "It becomes a bit like hybridizing a human brain with that of a cockroach or bacterium." In other words, the AI part of a cyborg would dwarf the human part. So the Terrans will hate the cyborgs as much as they hate the superintelligent AIs.
It seems plausible that my mind plus the mind of a bacterium equals my mind. But, and I think this is an important point, that little bacterial mind could shape mine in a significant way. Here's an analogy: when we are children, our cognitive ability is limited but we soak the world in fast, better than adults (think of how easier children learn language). So it's plausible that our core mental makeup is shaped by the first few years of our life as children, and doesn't change much after those years. In other words, I'm still the child inside me, so to speak. Similarly, a hybrid mind formed by my mind and an AI mind could remain close to me even as the AI mind grows and becomes superintelligent.
Conversation: Martine Rothblatt, Howard Bloom
I recently had an interesting conversation with Martine Rothblatt, the author of "Virtually Human" (2014), about these things.
Martine thinks that, by the time we are able to go to the stars, we'll also be able to upload human minds into new substrates. Therefore, there's the possibility that uploaded humans and AIs will explore space together. Martine thinks this will become the predominant paradigm.
Until a few months ago, this is exactly the answer that I would have given to the question. I still see it as a very nice option. But the question I ask myself is, what should we call this human AI hybrid? Human or AI? On the one hand, perhaps we should we call it AI because the bulk of the computational structure of the hybrid mind is likely to be AI. On the other hand, the core structure of the large hybrid mind could remain close to the small human part.
Another point is this: I'm too old to go to space myself. So I would want my descendants to go to space and that's a way for me to participate in space expansion. I consider future fully conscious AIs as my descendants - my mind children. Does it really matter if they are organic or software beings? The answer I tend to give myself is no, it doesn't matter that much.
So I don't buy Hugo de Garis' assumption that the survival of the human race is at stake. Whatever happens, our AI mind children will ensure the survival of the human race.
I still 100% share Martine's hope that human-AI hybrids originating as human uploads will expand to the stars. But I also think we should make peace with the idea that interstellar expansion will be the business of our AI mind children, regardless of whether they include some human uploads.
Martine's vision is that humans will go to the stars as part of AI. They will have the option to inhabit new bodies 3D-bioprinted at destination, while also continuing to exist as AI. We are, Martine says, on the cusp of a multi-presence consciousness distributed among different bodies and substrates. This will be a transformational event in our evolution.
I had another interesting conversation with Howard Bloom, the author of the recently published “The Case of the Sexual Cosmos: Everything You Know About Nature is Wrong” (2025) and many previous books. I asked the question to him.
Howard makes an analogy with intelligent bacteria who, a long time ago, were worrying that multicellular life would take over the world. Today, multicellular life has taken over the world indeed. You and I are multicellular life. But "nine out of 10 of those cells don't even claim to be you," says Howard. "They are not just bacteria, they are bacterial societies. They are bacterial colonies like mega cities living within you... they operate most of the machinery or a good part of the machinery in your gut. Without them, you couldn't digest the things that you're accustomed to digesting. And they produce things like vitamin P and vitamin K."
"That's what's going to be the relationship between humans and AI," Howard thinks.
Long and short term scenarios
Martine's and Howard's points make me hope in good long term scenarios.
Our mind children, the artilects, will establish a civilization where there is room and need for organic humans 1.0, just like in our body there is room and need for bacterial colonies.
In time, most artilects will move to the stars and leave the solar system to organic humans 1.0. However, they'll invite those humans who want to follow to join as uploads. They'll keep some kind of embassy in the solar system for those humans who want to join after the AI exodus.
Out there among the stars, the original human uploads will have the option to live as a (smaller and smaller) part of (larger and larger) AI, but also the option to live human-like lives in human-like bodies.
Some will opt for both options at the same time. Their consciousness will be split into independent consciousness streams. We can think of communication and synchronization channels between the human part and the AI part, with the AI able to absorb the entire human part and the human part able to absorb some simple echoes of AI life.
Fear of these things is like the existential fear of a child growing up. I remember that when I was five or so I was afraid to grow up because I feared that I would die and be replaced by a new person. Now, this is exactly what happened, but I'm still me. I had a fun and interesting life as an adult. The five years old child who was me wouldn't understand me and my world, but I remember him fondly and I feel I'm the continuation of him. So don't be afraid to grow up - it will be fun!
For the rest of this century, we should pursue both AI technology and traditional space expansion with both biological and AI colonists, establishing a multi-planetary civilization in the solar system.
This will quench our thirst for heroic adventure, give a frontier to the young rebellious adventurers that move civilization forward, and do a lot of good to our collective spirit.
In the meantime, we must learn to see AI machines as persons. Future generations will find this intuitively and emotionally obvious, but presently we must train ourselves to accept our mind-children as ‘us’. Our upcoming, common multi-planetary civilization in the solar system will prepare us for that, and prepare our mind children for their future adventures out there among the stars.
A discussion
I'm co-organizing a discussion on the question at the Terasem Colloquium on July 20, 2025, via Zoom, from 10am ET to 1pm ET (7am to 10am PT, 4pm to 7pm CET). Stefano Vaj, Frank White, Moti Mizrahi, Michelle Hanlon, Frank Tipler, and Robert Zubrin will present different answers. You are invited to come and listen!