What truly shapes a life? Why do some individuals thrive despite challenges, while others struggle even with apparent advantages? We know our genes play a role and our environment matters immensely, but the dynamic interplay between our biology, our experiences, and our choices has exceeded the understanding of both science and philosophy. We often receive generalized advice—eat well, exercise, manage stress—but how does your unique biology interact with your specific life experiences to determine your long-term health, wealth, and happiness?

I believe it's time to move beyond guesswork and generalized advice. So, The Human Trust is thrilled to announce “The Human Tapestry Initiative”, a groundbreaking project to build a revolutionary AI Foundation Model designed to decode the complex blueprint of human lives and tie personalized well-being to long-term health and human development.

Transformer models like ChatGPT learn to understand and generate human language with astonishing fluency by predicting the next best word and words given the many that have come before. Now, imagine an AI model trained not on words but on the deeply personal "language" of human lives—the dynamic interplay between our biology and our lived experiences. Rather than predicting the next word, such a model predicts the next experience, the next choice. That's the vision behind The Human Tapestry. This "moonshot" project (yes, despite my skepticism of “moonshots”, this is just such an unimaginably ambitious vision) will create the world's first Foundation Model trained on a unique combination of detailed biological data (focused initially on epigenetics, the molecular memory of our experiences, pair with other biomedical data) and rich behavioral information collected over time.

The Human Tapestry’s model will be powered by a one-of-a-kind Biobehavioral Data Engine. We will recruit a diverse group of 10,000 participants across the US and, with their full consent and data held securely within The Human Trust's nonprofit data trust, collect periodic biological samples for advanced epigenetic analysis. Crucially, we'll pair this with deep behavioral data gathered through our Human Trust platform that naturalistically captures daily activities, sleep, social interactions, and well-being. Additionally, we deploy a sophisticated AI chatbot (refined from our crisis intervention work) to gather rich, qualitative insights into experiences and challenges and the dynamics of individuals’ personality factors. Participants can also, if they choose, securely integrate data from their social media, economic, or educational lives, providing vital context.

This powerful combination of biological and behavioral information will train The Human Tapestry Foundation Model. Unlike current models focused on specific tasks, ours will learn the fundamental patterns that connect genetics, environment, behavior, and long-term outcomes. Its design not only predicts individual life trajectories—offering personalized insights into potential health risks (like depression, cancer, or Alzheimer's) or opportunities (in education or economic well-being)—but, even more importantly, to become a prescriptive tool. By understanding the intricate connections, the model aims to identify the specific, actionable behavioral or environmental changes that are most likely to positively alter an individual's predicted path today for better outcomes decades down the line.

The Human Tapestry is not just about building a powerful AI; it's about building a tool for precision public health and life trajectory mapping…and remapping. It's about moving from generic advice about “average people” to target guidance for proactively shaping our own life stories and lifting human capacity. This ambitious project is a testament to The Human Trust’s commitment to using AI for good, ensuring that this powerful technology serves to benefit individuals and build a future where everyone has the opportunity to thrive. We are building the map and the guide to help everyone navigate their unique journey.

Follow me on LinkedIn or join my growing Bluesky!

Research Roundup

Deviate!

In an AI-rich world, it is what makes you unique that makes you valuable. New research on developing brains suggests that being unique has always been valuable.

An ambitious analysis of “1,091 resting-state functional MRI scans of typically developing children from birth to 6 years of age” reveals that “individual deviations from normative [brain network] growth charts are significantly associated with infant cognitive abilities”. More idiosyncratic connectivity between “primary, default, control and attention networks were key predictors”.

As a mad scientist my first question is, can we intervene on functional network coupling? Can parents, peers, environment, and even neurotechnologies help every child develop into the deviants the world so desperately needs?

Life Choices? Choose Life.

Instead of obsessing over “job skills”, we should refocus policy and parenting on “life skills”. These don’t just lead to higher salaries—they produce longer lives.

Take Sweden. Over a 50-year period, “the gap in life expectancy between the top and bottom income percentiles widened substantially”: higher income men gained more than 7 extra years beyond their poorer peers, while the gap for women grew by nearly 5 years. This life expectancy gap increased in Sweden even as income inequality decreased and social spending expanded.

This growing gap has been driven by causes strongly tied to the social determinants of health: heart disease and cancer. In Sweden, money is buying better healthcare; rather, the life skills that afford greater income also substantially reduce the burden of “preventable causes” of death.

Quality Over Quantity

The UK Biobank is one of the great gifts to human development: a massive dataset of high-quality biomedical data on half a million participants. So much amazing research builds atop its foundation. How can we expand and extend the concept of biobaks?

A new study identified “2,663 biobanks and their textual mentions in 228,761 scientific articles, 16,210 grants, 15,469 patents, 1,769 clinical trials, and 9,468 public policy documents”. The highest impact biobanks have two defining characteristics. They are more “open to external researchers” and data quality, not quantity, “correlates with a higher impact in science, innovation, and disease”.

We need even more ambitious biobanks, and data banks across human development, structured as data trusts to bring even greater innovation into the world. Proprietary datasets may land venture funding, but true innovation is driven by open data.

<<Support my work: book a keynote or briefing!>>

Want to support my work but don't need a keynote from a mad scientist? Become a paid subscriber to this newsletter and recommend to friends!

SciFi, Fantasy, & Me

I just finished The Raven Scholar. A mash-up of murder mystery, young adult death match, and political thriller, the story does each of these with enjoyable flair (though not without the plot holes common to many “young adult death match stories”). Very good on its own, I loved the audio engineering of the audiobook: transition music, audio effects, and strong voice choices preserves fast perspective change style of the novel for listeners.

Stage & Screen

• May 23, Berkeley: I'll be talking "Responsible AI" at the Dutch Consulate.
• June 9, Philadelphia: "How to Robot-Proof Your Kids" with Big Brothers, Big Sisters!
• Late June, South Africa: Finally I can return. Are you in SA? Book me!
• September 18, Oakland: Reactive Conference
• October, UK: More med school education

If your company, university, or conference just happen to be in one of the above locations and want the "best keynote I've ever heard" (shockingly spoken by multiple audiences last year)?

Vivienne L'Ecuyer Ming