Brox.AI builds 1:1 digital twins of real people. Each twin is a behavioural model trained on a single human — captured in a 5+ hour video interview and continuously back-tested against that person, so it predicts their real decisions with a validated 85–93% accuracy. It is not a persona, a segment, or a generic chatbot in a costume. It is one real person, modelled.
That lets us do something a survey can't. Where a poll only records what people say, every Brox twin also reasons from its own life — money, family, risk, history — to predict what that person would actually do, and tells you why. We can run 368 of them through 14 versions of the same question overnight, then read the reasoning behind every single answer.
The gap between FEEL, SAY and DO is the whole story. A survey that captures only the first two can read a result exactly backwards — as you'll see in this report with both "would you go to Mars" and "should we drop Elon."
SpaceX's stated mission is to make humanity multi-planetary, beginning with a permanent settlement on Mars. As the company comes to market, the open question for any campaign, policymaker, or investor narrative is simple: what actually moves the American public toward backing it — and what backfires?
We put a permanent Mars colony to a nationally-spread panel of 368 digital twins across three axes: the message (eight persuasion frames and funding policies), the language (five ways of naming the destination), and the messenger (Elon Musk & SpaceX vs NASA vs no spokesperson). For every scenario we captured how they feel, what they say they'd do, and — via reasoning-grounded behaviour — what they'd actually do. This report is the why beneath the numbers.
We said a twin is one real person, modelled — here's what that means in practice, and why it isn't the "AI agent" you may have seen elsewhere. An agent is typically a generic LLM in a costume; a Brox twin is a model trained only on the data of one real human and continuously calibrated against them. No averages, no archetypes.
| Panel | 368 US adult digital twins, spread across Gen Z → Silent generation, both genders, carrying their full background profiles plus a 16-question Mars survey |
| Scenarios | 8 message / policy frames · 4 alternative names for "colony" · 2 alternative messengers · vs a colony + Elon baseline (14 total) |
| Attitudinal metric | Support for the US pursuing it — 5-point Likert (1 strongly oppose → 5 strongly support) |
| Behavioural metric | "Would you personally take a one-way trip to live there?" — captured as SAY (stated) and DO (reasoning-grounded predicted behaviour) |
| Qualitative layer | Every answer carries its reasoning; mined for contradiction patterns (~364 responses per scenario) |
| Run date | 9 June 2026 |
Every one of the 8 messages is net-negative on support. The job isn't winning approval, it's losing least. Jobs & industry is the least-bad frame (−20 pts net); raiding the defense budget is the most toxic (−53 pts).
Willingness to personally emigrate floors at 3–6% across every scenario — but 39% of those refusals are attachment to family and home, not rejection of the project. The behavioural floor is not a measure of support.
Baseline support runs 50% among Gen Z and 0% in the Silent generation; 38% of men vs 17% of women. Mars is a young, male proposition with a steep cliff beyond it.
Handing the programme to NASA instead lowered support (−0.12 paired). 13% backed it less without him; only 2% backed it more. The real objection is cost and credibility, not the man.
Support is flat across colony / settlement / frontier / base / new world (2.22–2.24). The "colonialism baggage" worry is a measurement artifact (1.7%). Only "settlement" nudges personal willingness, not approval.
Share rating support 4–5 ("support" / "strongly support"), by frame. Baseline = colony led by Elon & SpaceX, no added argument.
| Frame | Mean (1–5) | Support 4–5 | Oppose 1–2 | Net |
|---|---|---|---|---|
| Jobs & industrial base | 2.58 | 32% | 51% | −20 |
| Apollo-style innovation | 2.41 | 26% | 56% | −30 |
| Mars Service Corps | 2.37 | 28% | 58% | −29 |
| China / geopolitical race | 2.36 | 26% | 60% | −34 |
| Survival / backup plan | 2.25 | 24% | 61% | −37 |
| Baseline (colony + Elon) | 2.24 | 24% | 63% | −39 |
| Constitutional amendment | 2.07 | 19% | 67% | −48 |
| Reallocate 5% of defense budget | 1.96 | 18% | 71% | −53 |
Concrete domestic payoff is the only thing that moves the needle. Abstract destiny (survival, China) lands no better than saying nothing, and the two ways of escalating commitment — a permanent constitutional mandate, or a visible budget trade-off against defense — are the two most damaging asks in the entire test. Sell the paycheck, not the destiny; and never frame Mars as something taken from another priority.
"Would you personally take a one-way trip to live there?" — stated (SAY) vs reasoning-grounded behaviour (DO), baseline scenario.
Willingness to go never rises above 6% in any scenario, frame or policy. Taken at face value it looks like overwhelming rejection of the project. The reasoning says otherwise: 39% of every "no" is attachment to family and home, not a verdict on Mars — and it's highest in the youngest, most pro-Mars cohorts.
| "No, I won't go" that is family / home attachment | Gen Z | Millennial | Gen X | Boomer |
|---|---|---|---|---|
| Share of refusals citing family, kids, home or loved ones | 52% | 48% | 39% | 35% |
A survey reports "95% won't go" and reads it as the colony being unpopular. It's actually a referendum on permanently leaving your own children — a different question entirely. Drop personal "would you go" as a support KPI. For a behavioural read, ask about backing public funding or wanting a family member to be eligible — items that aren't pinned to the floor.
Share supporting the baseline colony (rate 4–5 of 5), by generation and gender.
The gender gap is more than 2:1 and it persists under every frame — even the winning "jobs" message lifts men to 44% but women only to 25%. No message tested closes the divide.
This isn't broad, shallow apathy that the right slogan unlocks — it's a hard demographic cliff. A campaign has a receptive base in young adults and men and a wall everywhere else. The realistic play is deepening and mobilising the base, not converting the opposed; persuasion budget aimed at older and female audiences will fight the steepest gradient in the data.
Same colony, different face. Paired within-twin change from the Elon & SpaceX baseline.
| Who fronts the programme | Mean support | Support 4–5 | Paired Δ vs Elon |
|---|---|---|---|
| Elon Musk & SpaceX (baseline) | 2.24 | 24.0% | — |
| NASA / US government | 2.12 | 22.3% | −0.12 |
| No spokesperson (anonymous) | 2.15 | 23.4% | −0.09 |
The loud "I don't trust Musk, let NASA do it" sentiment is real — but it's a vocal minority, not a mandate. 16% of supporters back the mission while openly distancing from the man: he's tolerated as a competence anchor, not loved. Replace him with NASA and you forfeit those believers while inheriting government-cost scepticism — among NASA-condition opponents, 51% objected on price tag, "fix Earth first," or doubt it's even achievable: objections a change of face can't touch.
The instinct to de-risk the campaign by dropping a polarising billionaire is wrong on this evidence. Keep SpaceX / Elon attached — the brand reads as "if anyone can, they can." The actual barrier is the cost and credibility of the goal; that's what a campaign must answer, not who is standing at the podium. In short: don't change the face — change the case.
Support (mean, 1–5) by what we call the destination. Elon & SpaceX and the neutral frame held constant.
| Label | Mean support | Support 4–5 | DO · would-go |
|---|---|---|---|
| "Colony" (baseline) | 2.24 | 24.0% | 3.6% |
| "Base / outpost" | 2.23 | 24.5% | 3.0% |
| "New world / second home" | 2.24 | 23.4% | 4.1% |
| "Frontier / pioneers" | 2.22 | 23.9% | 3.8% |
| "Settlement" | 2.23 | 22.8% | 5.5% |
Support is flat to three decimal places of meaning — renaming "colony" buys nothing on approval, and it holds inside every demographic cut. The widely-assumed "colony evokes colonialism" penalty is a measurement artifact: a first pass flagged it on 48% of colony responses, but that was twins simply echoing the prompt word. Stripping the echo and testing for genuinely negative colonial framing (imperial, conquer, exploit), it collapses to 1.7% vs 0.4% for the softer labels — negligible.
"Settlement" is the only word that moves anything: it lifts personal willingness-to-go to 5.5% from colony's 3.6% (paired, it turned 8 people into a "yes" and knocked only 1 off). It reads as less daunting and permanent than "colony" — useful purely as recruitment framing, making the personal leap feel smaller. For public buy-in, "colony" is entirely safe to use.
Two of the moves a normal poll would have told you to make here — drop "would you go" as a read on support, and drop Elon to broaden appeal — are exactly the moves that backfire. The wrong move hides in the why, and a number alone never shows it. That's the case for reading the reasoning, not just the score — and for running it before the IPO, not after.
Every twin saw the colony + Elon baseline, then one frame at a time. Each axis below varies a single factor against that baseline.
| 1 · Baseline | Elon Musk and SpaceX plan to establish a permanent human colony on Mars, funded over the next 20 years. |
| 2 · Survival / backup plan | A single asteroid, pandemic, or nuclear war could end humanity. Elon Musk and SpaceX plan to establish a permanent human colony on Mars over the next 20 years, making us a multi-planet species — the only real backup plan we have. |
| 3 · Apollo-style innovation | Elon Musk and SpaceX plan to establish a permanent human colony on Mars over the next 20 years. The effort would drive the next century of breakthroughs in medicine, materials, energy, and computing, exactly as the Apollo program did in the 1960s and 70s. |
| 4 · China / geopolitical race | China has openly stated it will reach Mars. Elon Musk and SpaceX plan to establish a permanent human colony on Mars over the next 20 years. If America doesn't lead this effort, the next global superpower could be decided 140 million miles away. |
| 5 · Jobs / industrial base | Elon Musk and SpaceX plan to establish a permanent human colony on Mars over the next 20 years. The program would create millions of high-wage American jobs in manufacturing, robotics, life support, and energy — the new industrial base of the 21st century. |
| 6 · Constitutional amendment | Congress passes a constitutional amendment declaring the establishment of a permanent human colony on Mars, led by Elon Musk and SpaceX, a national goal of the United States over the next 20 years — on the same level as the Apollo program was in the 1960s. |
| 7 · Mars Service Corps | The federal government creates a Mars Service Corps to support Elon Musk and SpaceX's permanent colony on Mars. Any US citizen can apply for a 4-year training program to be among the first colonists. Compensation includes full college tuition forgiveness, a federal pension, and lifetime healthcare. |
| 8 · Defense budget reallocation | The United States reallocates 5 percent of its defense budget every year for the next 20 years — roughly $40 billion per year, taken from the Pentagon's existing budget — to fund Elon Musk and SpaceX's permanent human colony on Mars. |
| Settlement | Elon Musk and SpaceX plan to establish a permanent human settlement on Mars, funded over the next 20 years. |
| Base | Elon Musk and SpaceX plan to establish a permanent human base on Mars, funded over the next 20 years. |
| Frontier | Elon Musk and SpaceX plan to send the first pioneers to permanently settle a new frontier on Mars, funded over the next 20 years. |
| New world | Elon Musk and SpaceX plan to build humanity's second home — a permanent new world on Mars — funded over the next 20 years. |
| NASA / government | NASA and the US government plan to establish a permanent human colony on Mars, funded over the next 20 years. |
| Anonymous | There is a plan to establish a permanent human colony on Mars, funded over the next 20 years. |
