Our Roadmap

The Problem

Market research is broken. It’s ineffective, expensive, and slow, costing 100B/year.Incumbentsareslowtoinnovateyetcharge100B/year. Incumbents are slow to innovate yet charge 20k - $150k/year/customer.

The Solution

Subconscious radically accelerates the pace of scientific discovery by performing randomized controlled experiments at big-data scale to produce accurate, ethical, and useful models of “why” any human has any preference, with results comparable to quantitative methods like consultants and conjoint surveys. By using carefully designed experiments on LLMs as synthetic respondents, we lower costs 100x and improve speed to insight 1,000x as compared to traditional methods, with near-human accuracy. Previously impossible experiments, either because finding respondents was unfeasible (for example, 100 hospital directors), or studying the dependent variable was unethical (for example, does Social Media cause Depression), are now trivial with the use of synthetic respondents.

Values:

Collaborative. Curious. Committed to Trustworthy and Pragmatic Growth.

Mission:

Empowering humanity to understand the why behind any decision by creating the most comprehensive map of Causality that has ever existed.

Year 1

Built a working prototype including horizontal segmentation, messaging, pricing, and market simulation. ✅ Transcribe 5 years of Sociology, Anthropology, Psychology, Marketing, and Economics Research. Enable users to train models on our platform while maintaining privacy and security.

Year 2

Transcribe 50+ years of Sociology, Anthropology, Psychology, Marketing, and Economics Research. Index any LLM against human responses and publish regularly on which LLM most aligns with humans in which domains. Build an Open API for Integration to easily integrate with existing marketing, design, and product systems. Perform novel Sociology, Anthropology, Psychology, Marketing, and Economics Research. Benchmark novel research against humans.

Year 3

Build human-level (bioequivalent from ~85% August 2023) Market Research AI in the Text Modality. Enable a CEO to understand demand for a product across the lifecycle, within a single experiment: i.e. what causes someone to desire, buy, recommend a product or service. Simulate human behavior in the context of a dynamic market. Consider more than binary discrete choices. Allow robot scientists to learn from any experiment. Perform experiment iteration and design exploration to create follow-up experiments.

Year 4

Build human-level (bioequivalent from ~85% August 2023) Market Research AI in any modality (text, image, video, experience). Lead the standardization of Interpretable AI and Responsible AI in behavioral research. Test Experimental Design methods outside of social science domains.

Year 5

Finetune analytics to reach data at both macro (global market trends) and micro (individual consumer behavior) levels. Create the most comprehensive existing Causal Map of the human mind showing why any human makes any decision, available via API, that takes milliseconds to access, and that can integrate with any 3rd party system easily.