Insight | Humans to Mars soon, robots first, or floating cities instead?

The Short Version

Five conversations on the Lex Fridman Podcast give five different answers to when people will live on Mars and why anyone should go. Jim Gates is certain no human will stand on Mars by 2030 and guesses at a date near the end of this century or later [2]. Martin Rees hopes for a small, uncomfortable community there by 2100 but argues that robots should do nearly all the useful work in the meantime [1]. Ariel Ekblaw and Adam Frank both expect most people in space to live in floating or spinning habitats rather than on Mars [3][4]. Rees, Ekblaw and Frank also share one conclusion: Mars is not a near-term escape from Earth's problems [1][3][4].

What They Said

Martin Rees: robots for the work, pioneers for the adventure

Martin Rees, introduced as an emeritus professor of cosmology and astrophysics at Cambridge, has written a book on exactly this question, The End of Astronauts, with Donald Goldsmith [1]. His argument starts with how quickly machines are improving. "Robots 50 years ago couldn't do anything very much," he says, but now they could assemble big structures in space or on the Moon [1]. The rovers on Mars today cannot really do geology, he concedes, yet "future AI will be able to do the geology," and they can already dig [1]. Humans, by contrast, need to be fed for the roughly 200-day journey out and brought home again, and robots need neither [1]. His conclusion is blunt: "So the practical case for humans is getting very, very weak" [1].

He backs that up with examples. Nobody is planning to send people to Europa or Enceladus, so exploring the outer solar system depends on robots anyway [1]. Cassini spent 13 years around Saturn on 1990s technology, and he asks listeners to think of how far smartphones have come since: "just think how much better one could do instrumenting some very small sophisticated probe" [1]. Dozens of such probes could explore the outer planets, he says, and he applies the same logic to Mars [1]. A large radio telescope on the far side of the Moon, shielded from Earth's artificial radio noise, could be assembled by robots without people [1]. "So on the moon and on Mars, I think everything that's useful can be done by machines, much more cheap than by humans," he says [1].

That leaves human spaceflight with one purpose. "And if humans go, it's only as an adventure, really," Rees says [1]. The line in his book, he explains, is that "human spaceflight should not be pursued by NASA or public funding agencies because it has no practical purpose" [1]. Part of his reasoning is about risk: public agencies have to be cautious, and the shuttle flew 135 times with two disasters that each killed seven people, one of them a schoolteacher [1]. He would leave human flights to billionaires and adventurers who accept high risks, calling it space adventure rather than tourism, and adds, "And we should cheer them on" [1].

Rees does expect people to reach Mars. He hopes that by the end of the century a small community will live there, perhaps on one-way tickets, "far less comfortably than at the South Pole or the bottom of the ocean or top of Everest" [1]. When Fridman asks whether he expects to see a human step onto Mars, he replies, "I think it's unlikely in my lifetime since I'm so ancient. But I think this essentially is going to happen" [1]. He doubts the public will pay for it, though: "I still think it's going to be hard to persuade the public to send people when you say straight out that they may never come back" [1].

What he rejects outright is Mars as a lifeboat. He calls it "a dangerous illusion" to think, as he says Elon Musk has suggested, that mass emigration can let people escape Earth's problems [1]. His reason is comparative difficulty: "it's far easier to deal with climate change on Earth than to terraform Mars" [1]. For ordinary people, in his words, "there's no planet B" [1].

His most unusual idea concerns who those pioneers will become. By the end of the century, he expects huge advances in genetics and in cyborg techniques, which will probably be heavily regulated on Earth on ethical grounds [1]. The settlers on Mars, beyond the regulators' reach, will have a reason to use them. "They are very poorly adapted to where they are. That's why they need this modification, whereas we're adapted to the Earth quite well," Rees says [1]. In his view these pioneers may become a new, post-human species, and since the Sun is less than halfway through its life, humans are unlikely to be the end point of evolution [1].

S. James Gates Jr.: not by 2030, and the body is the hard part

19. James Gates Jr., a theoretical physicist and science fiction fan, frames the question through money and biology [2]. When Fridman asks whether a human will reach Mars, he does not hedge: "I am extraordinarily convinced we will not put a human on Mars by 2030" [2]. He notes that 2030 is a date often heard in the public debate [2].

His explanation uses the Moon landing as the yardstick. About ten years passed between President Kennedy's challenge and the landing, and NASA, he says, "consumed about 5% of the country's economic output" over that period to travel 250,000 miles [2]. "Mars is about 100 times farther," he continues, and the government is spending about a tenth of what it spent then [2]. Putting those together, he says, "it's at least 1,000 times harder for me to imagine us getting to Mars by 2030" [2].

Money is the easiest part to explain, but biology worries him most. Earth's magnetosphere shields us from cosmic radiation, and he cites estimates that a traveller sent to Mars with current technology would spend about two years outside that shield [2]. "That means radiation probably means cancer," he says [2]. The real challenge, as he frames it, is how to engineer a ship or a body so that someone recognizably human can "survive the rigors of interplanetary space travel" [2]. "It's much more difficult than most people seem to take into account," he adds [2].

Still, he is willing to go out on a limb. He suggests a human may reach Mars somewhere around 2090 to 2120 [2]. When Fridman raises Elon Musk and Jeff Bezos, Gates agrees that "bending the cost curve is certainly critical if we're going to be successful" [2]. But he says he is disappointed by the routes they have chosen [2]. Rockets landing on their fins are astounding to watch, he says, yet the technology underneath is not very different from what went to the Moon [2]. He describes an elongated engine design that lets thrust be varied during ascent and has had preliminary testing, and wonders why entrepreneurs are not pursuing it; what he sees instead is incremental change [2].

Gates also offers a way around the body problem entirely. He does not believe biological wetware has an exclusive patent on consciousness, describing humans as "self-actuating, self-learning data streams" [2]. If that kind of data stream could be exported to hardware, he says, "that kind of consciousness could get to Mars very quickly. It doesn't have our problems" [2]. Sending a mind at the speed of light is harder still, he adds, because it would require preserving the information far beyond how the internet packages messages today [2].

Ariel Ekblaw: outposts on Mars, cities in orbit

Ariel Ekblaw, founder and director of the MIT Space Exploration Initiative, builds the kind of structures she thinks people will actually live in [3]. She admires the effort to reach Mars but says the planet is misunderstood as a near-term backup for Earth [3]. Her list of reasons is practical. The soil contains toxic perchlorates, the atmosphere is too thin, and a settlement would have no supply chain for advanced technology: if your electronics break, "there's no radio shack" [3].

Her timeline is earlier than Gates's for a first step, but modest about what that step would be. She thinks small outposts, similar to McMurdo station in Antarctica, are feasible in the 2030s, with first crewed missions possible around then [3]. Life there, she says, would be like pioneer days [3]. A Mars round trip takes about three years, including six to nine months each way and a wait for the planets to line up again [3].

She names two central challenges for missions that long. The big one is radiation: beyond the magnetosphere, crews face higher cancer risks later in life, and heavy shielding such as lead adds a great deal of mass [3]. The one that seems small but is actually huge is mental health, keeping crews happy, balanced and cooperating [3]. She adds that NASA still has not settled whether married couples or single crew members make better teams, and she points to Scott Kelly's year in space and the twin study, which documented changes in bone density, muscle and eyesight [3]. For missions beyond about three years, she believes spinning habitats with artificial gravity will be needed, and she suspects that pregnancy and early childhood in space will require a 1G environment [3].

Then comes what she calls her hot take: larger human populations in space will live in floating microgravity cities [3]. They would be built piece by piece through self-assembly, not through a single trillion-dollar investment [3]. She expects them to grow in pockets near resources, such as the Moon, asteroids and gravitationally stable Lagrange points, while acknowledging that crowding low Earth orbit worries astronomers [3].

Her own research is the building block for those cities. Her PhD project, Tesserae, starts from the problem that the structures people like Jeff Bezos imagine for millions living in space are bigger than any rocket's payload fairing [3]. Her answer is flat-packed tiles that float free in microgravity and pull themselves together with electro-permanent magnets on their edges, with no human in the loop and no central coordinator [3]. The tiles carry sensors that detect a bad bond so they can let go and try again, which she compares to error correction in DNA [3]. She has tested three-inch tiles on the International Space Station, is building tiles about 2.5 feet on each edge, and aims for a buckyball roughly 10 meters across, far roomier than any current station module [3].

She favours decentralized designs for safety. Losing a few swarm robots does not matter, and a station made of separate nodes lets a crew seal off one section during a fire or conflict and move elsewhere [3]. Her dream structures include a space cathedral with long sight lines, a station of space-filling nodes that could grow into a nautilus shape following a Fibonacci spiral, and a ring world circling a planet or moon [3].

Ekblaw's reasons for going are also different from Rees's adventure. Her military family's motto was "luck favors the prepared," but rather than treating space as an escape hatch, she finds it more compelling to use space technology to keep Earth livable, through climate-monitoring satellites and possibly space-based geoengineering [3]. Building habitats forces innovations that come back home, she says, such as energy-efficient cooling, air filtration, carbon dioxide scrubbing and resilience to extreme conditions [3]. She gives the example of Californians who now need nearly airtight homes during wildfires [3].

Adam Frank: centuries for Mars, a solar system of habitats

Adam Frank, an astrophysicist who studies how star systems evolve, sees the long-term picture as large and the near-term picture as sobering [4]. He expects the next thousand years of human history to be the solar system, provided humanity survives climate change, nuclear war and AI [4]. "Within a few hundred years, it is not unimaginable that there will be millions, if not billions of people living in the solar system," he says [4]. But he adds that "most of them will be in space habitats versus on Mars on the planetary surface" [4].

His habitat idea came from testing a science fiction premise. In the TV series The Expanse, people hollow out asteroids and spin them for gravity, and Frank's group checked whether that could work [4]. Sadly, he says, it cannot: spun up to the speed needed for one-third gravity, "which is what I think the minimum you need for human beings," the rock "would just fall apart" [4]. The group came up with another idea instead. Wrap a small asteroid, a couple of kilometers wide, in a giant nanofiber bag and spin it, and the bag inflates until "you could get like a Manhattan's worth of material inside" [4]. He pictures many such habitats as experiments in "maximum flourishing, maximum wellness, maximum democracy, maximum freedom" [4]. He does not expect harmony, though: drawing on The Expanse, he says "there will be warfare and there will be conflict" [4].

On Mars as a backup plan, Frank is cautious. The threats of climate change and nuclear war, he says, are "breathing down our necks now" [4]. A base on Mars will not be close to self-sufficient for a century at least, so "it's not like a backup plan now" [4]. Self-sufficiency would require a whole material economy that he puts two to three centuries away [4]. He puts that span in perspective: "look at 1820, nobody had traveled faster than 60 miles an hour unless they were falling off a cliff," and now people routinely travel at 500 miles an hour [4]. His conclusion is that "the odds that we're gonna be able to create a self-sufficient colony on Mars before that threat comes to head is small," so the threat itself must be faced [4].

When Fridman argues that a looming catastrophe could speed the effort up, Frank agrees it is a fascinating engineering question but still holds that the problems must be solved here [4]. He also sees a useful link between the two. Building a truly self-sufficient habitat means building something as robust as an ecosystem, which could teach people how to live sustainably on Earth: "I think the two problems are going to kind of feed back on each other" [4]. He illustrates the stakes with Kim Stanley Robinson's novel Aurora, in which colonists find a planet that seems habitable but proves deadly to Earth life [4]. For Frank, the biosphere is "the true unit of life" [4]. If humanity gets through climate change, he likes to say, "the prize at the end is the solar system" [4].

Frank also warns against assuming technology will keep racing ahead. We live in a great acceleration, he says, and may wrongly assume technology always advances; he jokes that in Assassin's Creed, 300 years pass and everyone is still stabbing each other, which was true for most of human history [4]. Faster-than-light proposals such as the Alcubierre warp drive rely on exotic matter that, he says, does not exist [4]. He grants that a breakthrough in physics could change everything, but notes that string theory has had 40 years of effort [4].

Brian Greene: engineering first, physics later

Brian Greene, a theoretical physicist at Columbia, comes at the question from the angle of what physics allows [5]. Asked whether colonizing space is a physics or an engineering problem, he answers, "I think it's fundamentally an engineering problem" [5]. Without wormholes or near-light-speed travel, he says, colonization will proceed very slowly [5]. Relativity offers a striking loophole: travel close enough to the speed of light, and "you can actually go arbitrarily far in a human lifetime" [5]. He mentions solar sails as a vital and valuable first step [5].

He then refines his answer. "I should probably modify my answer from before," he says; it is "really both" [5]. In his view, "we will surmount the engineering challenges and that will then make the physics challenges relevant," such as learning to travel near light speed or to manipulate the shape of spacetime [5]. He calls it "a multi stage process where it is engineering and ultimately physics" [5]. The biggest obstacle he names is not technical. Raising the possibility that advanced civilizations destroy themselves, he says, "that's the big obstacle" [5].

Greene sees exploration as core to the human spirit and thinks a child could be born on Mars within Fridman's lifetime [5]. Discussing travel to the future near light speed, he lists the costs plainly: "There are engineering challenges, there are technological challenges. And there are even biological challenges, right? There are G forces that you're going to experience" [5]. Those, he says, are the details; the universe allows the journey [5]. Fridman, who was born in the Soviet Union, says he would take a one-way trip and is moved by the idea that someone born on Mars could look at Earth through a telescope "and say that's where I came from" [5].

Fridman's pushback

Fridman hosts all five conversations, and on this question he argues back with Rees and with Frank. With Rees, he says Musk's aim is a small backup plan rather than an escape, and that becoming multi-planetary is a long-term investment that inspires young scientists and drives discoveries [1]. He calls reaching into the unknown "one of the most inspiring things humans can do" [1]. He also argues that the negative framing of billionaires misses the real difference between companies and governments [1]. With Frank, he suggests that a looming catastrophe could hasten a Mars colony [4].

Where They Agree, and Where They Don't

Agreement: Mars is not a lifeboat for now

Rees, Ekblaw and Frank each reject the idea of Mars as a near-term escape. Rees calls mass emigration "a dangerous illusion" because fixing the climate here is easier than terraforming Mars [1]. Ekblaw calls the planet misunderstood as a near-term backup and prefers to use space technology to keep Earth livable [3]. Frank says a Mars base is not a backup plan now because self-sufficiency is centuries away and the threats are immediate [4]. Fridman is the dissenting voice, describing Musk's goal as a modest backup and a source of inspiration [1].

Agreement: the human body is the bottleneck

Gates and Ekblaw both put radiation at the top of the list. Gates says two years outside the magnetosphere probably means cancer [2], and Ekblaw says crews face higher cancer risks later in life while shielding adds heavy mass [3]. Ekblaw adds mental health, bone, muscle and eyesight, and the need for artificial gravity on longer missions [3]. Frank sets one-third gravity as the minimum he thinks humans need [4]. Rees and Gates both reach for the same unusual answer, changing the human rather than the environment: Rees through genetic and cyborg modification of the Mars pioneers [1], Gates through engineering the body or exporting the mind to hardware [2].

Disagreement: the timeline

The dates spread widely. Gates is "extraordinarily convinced" there will be no human on Mars by 2030 and guesses roughly 2090 to 2120 [2]. Ekblaw thinks first crewed missions and McMurdo-style outposts are possible in the 2030s [3]. Rees expects a landing to happen, though not in his lifetime, and hopes for a small community by 2100 [1]. Greene thinks a child could be born on Mars within Fridman's lifetime [5]. Frank puts a self-sufficient colony two to three centuries out [4].

Some of this gap is about what is being measured. Ekblaw's 2030s outpost is a small, supplied base like an Antarctic station [3], while Frank's centuries are for a colony that can sustain itself without Earth [4]. Gates bases his estimate on funding compared with Apollo and on the radiation problem [2]. Rees's caveat is political: he thinks the public will be hard to persuade to send people who may never return [1].

Disagreement: who should send people, and why

Rees says NASA and public agencies should not fund human spaceflight at all, since it has no practical purpose [1]. Ekblaw sees the shift differently: NASA is not losing ground, she says, but freeing itself from maintaining a station in low Earth orbit so it can push toward Mars and Europa [3]. She praises SpaceX's willingness to fail publicly, something a taxpayer-funded agency could not do [3].

The reasons for sending people differ as much as the dates. For Rees, the reason is adventure [1]. For Ekblaw, it is technology that keeps Earth livable and a new kind of society in orbit [3]. For Frank, it is experiments in flourishing and lessons in sustainability that feed back to Earth [4]. For Greene, exploration is part of the human spirit [5]. For Fridman, it is inspiration for the next generation [1].

Disagreement: are today's rockets a breakthrough?

Rees calls SpaceX's reusable rockets a tremendous service, making giant telescopes and even space-based solar power feasible [1]. He credits a Silicon Valley culture of younger people, contrasting it with big aerospace companies whose staff were in some cases "left over from the Apollo program" [1]. Gates agrees that cost matters but is disappointed: landings on fins are astounding, yet he sees the technology as not very different from Apollo's and the change as incremental [2]. Ekblaw is thrilled by commercial spaceflight, which lets her buy access through SpaceX, Nanoracks and Blue Origin instead of waiting years for grants [3].

Disagreement: Mars or habitats?

Here the five split into camps. Ekblaw and Frank both expect most people in space to live in free-floating or spinning habitats rather than on Mars [3][4]. Rees expects a small, hardy community on Mars and leaves big construction to robots [1]. Greene and Fridman speak of Mars colonization and a child born there [5]. Gates's concern is more basic: getting a living, recognizably human traveller there at all [2].

Why It Works

The sources give several mechanisms behind these positions. The first is cost and payload. Rees's case rests on robots needing neither food for the roughly 200-day journey nor a trip back home, which makes them far cheaper [1]. Gates's estimate rests on the ratio of distance to spending compared with Apollo [2]. Ekblaw's self-assembling tiles exist because large structures will not fit in a rocket fairing, and her buckyball shape was chosen because it gives the most volume for the least surface material, which is what is expensive to launch [3].

The second is the environment beyond Earth's magnetic shield. Gates and Ekblaw both describe the magnetosphere as the protection crews lose on the way to Mars, with cancer risk the result [2][3]. Microgravity weakens bones and muscles and changes eyesight, as Ekblaw's account of the Scott Kelly study shows [3]. Spin can stand in for gravity, but Frank's group found that bare rock cannot be spun fast enough for one-third gravity without breaking apart, which is why they turned to a nanofiber bag [4].

The third is self-sufficiency. Ekblaw's "there's no radio shack" and Frank's material economy describe the same problem: a settlement depends on supply chains it cannot yet build for itself [3][4]. Frank links this to the biosphere, arguing that a self-sufficient habitat must be as robust as an ecosystem [4]. Ekblaw adds that decentralized designs give redundancy, so the loss of one part does not doom the whole [3].

The fourth is how technology advances. Rees says progress in a technology "goes up very fast and then it levels off," citing Apollo, only 12 years after Sputnik, which "was and still remains the high point of human space exploration" [1]. He adds jumbo jets that have barely changed in 50 years [1]. Frank warns that living through a great acceleration biases people toward assuming progress is automatic [4]. Greene's staged picture, engineering first and physics later, is the more hopeful version of the same idea: each stage makes the next one's problems relevant [5].

Where It Applies, and Where It Doesn't

The robots-first argument applies most clearly to science and construction. Rees's examples, including probes to the outer planets, robotic geology on Mars and a radio telescope on the far side of the Moon, are all jobs where a machine can do the useful work [1]. Ekblaw's swarm robots that crawl over spacecraft looking for punctures and leaks fit the same pattern, and her lab's AstroAnt was set to ride a rover to the lunar south pole [3]. Rees's argument does not apply, by his own account, to people who want to go for the adventure, and he would cheer them on [1].

The pessimistic timelines apply to self-sustaining settlement and to government-funded crewed missions. Gates's 2030 doubt rests on public spending and radiation [2], and Frank's centuries rest on building a full material economy [4]. They do not rule out a small supplied outpost, which Ekblaw thinks is possible in the 2030s [3]. They also leave room for change: Gates says bending the cost curve is critical [2], and Frank grants that a breakthrough in physics could change everything [4].

The floating-city vision applies to large populations over the long run. Ekblaw sees such cities growing piece by piece near resources [3], and Frank sees millions or billions in habitats within a few hundred years [4]. It does not yet apply to anything people can live in: Ekblaw's tiles are still being scaled up from three inches to 2.5 feet, and her 10-meter buckyball is a goal [3]. Frank's bagged asteroids are an idea from a paper [4].

The "no planet B" view applies to the question of escaping climate change and nuclear war in the coming decades [1][3][4]. It says less about the far future, where Rees, Frank and Greene all picture humans or their descendants spreading out [1][4][5].

What To Do With It

1. Ask what a Mars date measures. When you hear a target such as 2030, check whether it means a single landing, a supplied outpost or a self-sustaining colony; Gates, Ekblaw and Frank give very different dates because they mean different things [2][3][4].
2. Use the Apollo yardstick. Gates compares distance and spending with the Moon program, about 5% of economic output for ten years, to judge whether a timeline is realistic [2]. Apply the same test to any announced plan.
3. Watch radiation and biology, not only rockets. Gates says surviving interplanetary travel is much harder than most people take into account [2]. Ekblaw calls radiation the big one among the challenges of long missions [3]. Progress on shielding, artificial gravity and crew health is a better sign of real readiness than a new launch vehicle.
4. Judge cost breakthroughs by what changes underneath. Rees credits reusable rockets with making giant telescopes feasible [1], while Gates looks for new engine designs rather than incremental change [2]. Ask which kind a new announcement is.
5. Count robotic missions as exploration. Rees argues that probes and robots can do what is useful on the Moon and Mars far more cheaply [1]. Follow robotic missions as the main way exploration will advance in the near term.
6. Look to orbit, not only to Mars. Ekblaw and Frank both expect most people in space to live in habitats [3][4]. Self-assembling structures, commercial stations and spinning habitats are the projects that test their view.
7. Treat space as a tool for Earth. Ekblaw points to climate-monitoring satellites and habitat technologies such as air filtration and carbon dioxide scrubbing that come back home [3], and Frank expects habitat design to teach sustainable living [4].
8. Keep the Earth's problems first. Rees, Ekblaw and Frank each argue that climate change and other threats must be solved here, because Mars cannot be made into a refuge in time [1][3][4].
9. Expect people to change as well as machines. Rees thinks Mars pioneers will modify themselves with genetic and cyborg techniques [1], and Gates imagines exporting minds to hardware [2]. Debates about human enhancement are part of the Mars question, not separate from it.

Sources

1. Martin Rees on Dark Matter, Alien Life, Black Holes, Mars and the Risks That Could End Us · Lex Fridman Podcast
2. Jim Gates on Supersymmetry, String Theory, Hidden Codes in Physics and Proving Einstein Right · Lex Fridman Podcast
3. Ariel Ekblaw on Self-Assembling Space Habitats, Floating Cities, and Life Beyond Earth · Lex Fridman Podcast
4. Alien Civilizations and the Search for Extraterrestrial Life · Lex Fridman Podcast
5. Brian Greene on Entropy, Meaning, String Theory, Time Travel and the Fear of Death · Lex Fridman Podcast