- Startup Brox argues that traditional market research is too slow in the AI era, where a TikTok trend can shift global consumer behavior in mere hours.
- The company built a system of 60,000 “digital twins”—one-to-one AI replicas of real humans with full demographic, psychological, and consumer preference profiles.
- CEO Hamish Brocklebank asserts that these are not “synthetic humans” randomly generated by AI, but behavioral clones based on real data collected from participants with full consent.
- Brox recruits real people like a traditional survey firm, paying for in-depth interviews and collecting extremely detailed personal data.
- Some digital twins contain up to 300 pages of text data regarding childhood, relationships, marital status, and individual decision-making motivations.
- The company claims that “synthetic audiences” created with Large Language Models often produce “AI slop” because responses are over-converged and biased toward “proper” behaviors like healthy eating.
- Brox’s reasoning chain system allows the AI to explain step-by-step why a digital twin made a specific decision, helping businesses understand both “what happened” and “why it happened.”
- Clients can ask extremely sensitive questions, such as how bank depositors would react if the US invaded Iran or Greenland, or the impact of RFK Jr.’s statements on vaccine attitudes.
- Brox currently operates in the US, UK, Japan, and Turkey, with high-value data groups like individuals with assets over $5 million or dermatologists in rare specialties.
- To keep data up to date, the company continuously reconnects with the real people behind the digital twins. Some wealthy participants are granted Stock Appreciation Rights instead of cash to motivate them to maintain high-quality data.
- Brox focuses on two main sectors—finance and pharmaceuticals—for example, simulating bank run waves or predicting vaccine hesitancy amid political volatility.
- Brox’s business model is high-end SaaS with prices starting at $100,000/year and reaching up to $1,500,000/year for global deployment.
- Customers are allowed unlimited simulations during the contract period to encourage “testing everything” before making real-world decisions.
- Brox states they dislike prediction market models like Polymarket or Kalshi because betting platforms only predict outcomes without explaining the psychological motives behind human behavior.
- The company is expanding to the Middle East and APAC with the long-term goal of simulating the entire world into a “parallel universe” for risk-free decision-making.
📌 Conclusion: Brox has an immense ambition to replace traditional market research with “digital twins” of real humans. With 60,000 digital twins containing deep psychological and behavioral data, businesses can run surveys and simulate market reactions in hours instead of months. The biggest differentiator is that Brox avoids synthetic AI personas in favor of AI replicas of real individuals to prevent “AI slop.” Currently pricing its services between $100,000 and $1.5 million per year, the company aims to create a “parallel universe” simulating all of human society for businesses to test decisions before real-world implementation.
