Episode Summary
Executive Summary: Peter Diamandis interviews Dr. Mary Lou Jepsen about Open Water, a company using miniaturized optics, ultrasound, AI, and chip technology to build low-cost diagnostic and therapeutic devices for stroke, cancer, depression, addiction, and more. Jepsen frames the mission as open-sourcing a medical platform to democratize biology and accelerate approvals, data, and impact.
Main Topics: Open Water’s mission and platform (Priority: 5/5): Jepsen explains Open Water as a convergent-technology company using physics, AI, chips, lasers, ultrasound, and imaging to diagnose and treat disease at low cost, with the goal of making advanced medical tools widely accessible. Stroke detection and triage (Priority: 5/5): The company’s imaging system uses smartphone-era camera chips and lasers to detect large vessel occlusion strokes quickly enough to route patients to the right hospital and treatment center. Therapeutic ultrasound for brain and disease states (Priority: 5/5): Jepsen describes ultrasound modules intended to selectively target cancer cells, overfiring neurons, microclots, senescent cells, and possibly addiction or depression-related circuits without harming healthy tissue. Open-source business model (Priority: 4/5): A major theme is Open Water’s decision to open-source all patents, software, and hardware after a large grant/investment from Vitalik Buterin, aiming to reduce costs, increase trust, and speed adoption. Jepsen’s personal journey and moonshot mindset (Priority: 4/5): Her childhood tinkering, brain tumor survival, art background, and prior roles at Intel, Google, Facebook, Oculus, and One Laptop per Child shaped her bias toward ambitious, socially meaningful engineering. Healthcare system failure and regulatory bottlenecks (Priority: 5/5): The conversation emphasizes how slow, expensive clinical development and fragmented reimbursement keep promising medical technologies from reaching patients, motivating a new platform-based model.
Key Arguments: Physics plus Moore’s Law can shrink room-sized medical devices into handheld, low-cost platforms that both diagnose and treat disease. The same core hardware can target multiple diseases, making healthcare innovation more data-rich, safer, and economically scalable than single-disease drug development. Open-sourcing medical hardware/software can function as a trust model and distribution strategy, allowing many labs and countries to adopt and improve the platform. The current drug/device model is too slow and expensive for rapidly evolving medical needs; a reusable platform can reduce cost and time to impact. Selective resonance and focused ultrasound may affect diseased cells or neural circuits without damaging surrounding healthy tissue. More distributed testing and real-world usage can generate the data needed for regulatory approval faster than traditional small clinical trials. Jepsen argues that the future of healthcare will resemble consumer electronics: iterative, modular, lower-cost, and widely available. Her personal history of illness and recovery influenced her determination to build tools that prevent other people from suffering delayed diagnosis or inadequate treatment.
Data Points: Global annual deaths: 55 million - Peter says the world loses roughly this many people each year to disease during the discussion of healthcare urgency. U.S. healthcare spend share: About 25% of the U.S. economy - Peter cites the scale of healthcare spending to underscore system inefficiency. Hospital spending share: 28% - Peter cites this as part of the healthcare cost structure. Doctor spending share: 20% - Peter cites this as another major segment of healthcare expense. New drug approval cost: About 26 years and nearly $3 billion capitalized cost - Jepsen and Peter discuss how long and expensive it is to bring a new drug to market. Novel medical device approval cost: About $700 million and 13 years capitalized cost - Used to explain the regulatory burden for device innovation. Average U.S. new device approval cost: $658 million - Jepsen cites this as the average capitalized cost for regulatory approval over the last 30 years. Regulatory cost share tied to device development: 85% - Jepsen says most cost comes from building the device, not the trials themselves. Clinical trial cost per patient: $40,000 to $70,000 - Used to illustrate why large trials are prohibitive. Hospital availability for stroke procedure: 5% - Peter notes only 5% of U.S. hospitals can perform the needed thrombectomy procedure. Stroke treatment window: About 2 hours - Jepsen explains the narrow time window for large vessel occlusion stroke intervention. Glioblastoma test dose: 2 minutes - She describes the successful mouse work using a short ultrasound dose. Ultrasound diagnostic setting: Lower than used on pregnant women and fetuses - Jepsen uses this to emphasize safety levels in the glioblastoma work. Ultrasound frequency used in one experiment: 150 kilohertz - The best-performing glioblastoma treatment in mouse studies. Depression study size: 20 people - She describes an early clinical study at the University of Arizona. Depression remission: Nearly half - Jepsen says nearly half of severe depression patients went into remission in that preliminary study. Stroke imaging study size: 151 patients - She cites early results from Pan and Brown in the cath lab. Open-source grant: $50 million - Vitalik Buterin’s support was conditioned on Open Water open-sourcing its technology. Current device price: $10,000 - Jepsen says the system is being sold at this price now. Future device price target: Around smartphone cost / $1,000 range - She says volume manufacturing could bring it down dramatically. Battery cycle life: 2,000 charge-recharge cycles - She cites the durability innovation in the One Laptop per Child battery design. One Laptop per Child output: Millions - Jepsen says millions of laptops were produced.
Pivotal Quotes: "These are converging exponentials. This is the intersection of physics and AI and chip sets." — Peter Diamandis: Peter summarizes the technological convergence he sees in Open Water’s approach. "We can see blood flow 20 times better than a multi-million dollar MRI machine or CT machine or anything else we can find published in the literature." — Mary Lou Jepsen: She describes the performance of the miniaturized imaging system built from smartphone camera chips and lasers. "All 68 of our patents, all of our software, all of our hardware. Open source AGPL." — Mary Lou Jepsen: Jepsen explains the decision to make the platform fully open source after Vitalik Buterin’s backing.
Implications: If Open Water’s platform works at scale, diagnostics and therapies could become cheaper, faster, and more distributed, shifting medicine toward open, data-rich, consumer-electronics-like innovation and away from slow, siloed, drug-centric development.