Back to All-In Podcast

Daniel Ek: Life After Spotify, Broken Healthcare Incentives, Catching Disease Early & AI's Potential

All-In PodcastSeptember 28, 202651m
In a Nutshell

Daniel Ek built Neko Health to fix healthcare's broken incentive structure—where insurers and employers avoid investing in prevention due to short employment tenures and 20-year ROI timelines—by offering $499 annual comprehensive assessments that use multimodal data and AI to catch chronic diseases early when they're cheap to treat. The company has found serious undiagnosed conditions in 1% of 100,000 scans and demonstrates that vertically integrated diagnostics can generate positive unit economics while shifting from reactive to preventative care. Ek argues the healthcare system needs 10-100x more longitudinal data to enable meaningful population-level insights, and advocates for open-source AI to prevent capability concentration among few entities.

AI-Generated Notes

These notes were generated by AI and may contain inaccuracies.

Daniel Ek, co-founder of Spotify, started the company in 2006 at age 23. At that time, the music industry was experiencing significant losses due to widespread music piracy through platforms like Napster and Kazaa. The RIAA was actively suing individual consumers in the United States for illegal music downloads.

Ek and co-founder Martin were brainstorming startup ideas when Ek expressed his passion for music but initially dismissed it as a bad idea due to the complexities of licensing, copyright, and record company deals. Through a series of "what if" questions from Martin, Ek eventually committed to the concept of providing access to all the world's music at one's fingertips.

Sweden was identified as one of the worst music markets globally due to its super fast broadband connections from the early 2000s, which enabled rampant illegal downloading. iTunes was not even available in Sweden, leaving literally no legal options for consumers.

Ek and Martin used their own money from previous startups to approach record labels with a guarantee that they would maintain their bonus and budget for the next year. If the service failed after a year, the labels would still receive their bonuses; if it succeeded, everyone would benefit. The labels agreed after several years of negotiation.

Spotify launched in late 2008 and became an immediate success. It expanded to the UK and then launched in the United States in 2011. The company reached over 700 million active users and over 300 million premium subscribers.

Ek shifted from Spotify's CEO role to executive chairman on January 1st of the current year. He had been thinking about healthcare since at least 2012-2013, shortly after Spotify's US launch. He identified healthcare as a major area where spending more money was resulting in worse outcomes.

Ek became a Spotify shareholder in 2018 when the company went public, after which he began angel investing. However, he concluded that he preferred building companies over investing, leading him to focus on Neko. He noted that watching companies he invested in get mismanaged was painful, and that providing instruction to portfolio companies often led to failure.

Neko was founded in 2018 in Sweden. It took five years to develop the product, launching in Sweden in 2023 before expanding to the UK and now the United States. The US spends 18% of its GDP on healthcare, with heart disease alone costing hundreds of billions of dollars annually - larger than Fortune 10 company revenues for just one disease category.

Ek identified chronic disease as the major healthcare problem, noting that early discovery makes these conditions totally preventable with minimal cost and suffering. The key insight is that the healthcare industry lacks sufficient data to shift from reactive to preventative care.

Neko's solution centers on gathering more comprehensive data across multiple modalities and longitudinally over time. This approach mirrors Spotify's success with data-driven music recommendations. The company leverages smartphone sensors and AI (previously called machine learning) to make predictions from massive datasets.

Neko offers a $499 service that provides a reinvented healthcare experience through vertical integration. The company builds its own facilities, employs its own nurses and doctors, creates its own diagnostic equipment, and develops its own software. The service is designed to create "the most valuable hour you can invest in your health."

During a typical one-hour visit, patients receive:

  • Blood work measuring 53 blood markers
  • Skin analysis using a camera system taking over 6,000 high-resolution images to index every mole, lesion, rash, and redness
  • Heart and blood circulation examination
  • Grip strength measurement

Patients receive uninterrupted time with a clinician to review all results and discuss potential health improvements.

Neko has conducted over 100,000 scans across eight years of operation. Their third-year data survey shows that approximately 1% of members have serious undiagnosed medical conditions discovered through the service. The company publishes health outcomes data annually.

Notably, members with the worst initial health status show the most improvement. Examples include members who stopped smoking after their Neko assessment and others who received treatment for previously undiagnosed serious conditions.

The average Neko member has 950 moles. The company uses AI to flag potential risk factors across all moles, followed by human clinician review and expert dermatologist consultation when needed. Longitudinal data allows tracking of mole changes year-over-year, which individual doctors cannot reasonably remember.

Ek serves as co-founder alongside Yalmer, who is described as the "brainchild" of Neko and handles most operational work. This mirrors Ek's current Spotify arrangement with two CEOs.

Ek recommends annual Neko assessments, comparing this to yearly dental visits. While $500 represents significant cost for many people, he views it as a worthwhile health investment. The company recommends members do assessments on an annual basis.

The transcript cuts off mid-sentence while discussing the economic model of being vertically integrated and cutting costs dramatically.

Neko Health operates within a $500 annual subscription model where unit economics are positive. The company has clinics that are already profitable today. The pricing strategy positions the service as a great value for customers while remaining economically viable for the business, enabling growth and expansion to additional locations.

Daniel Ek is not exclusively focused on Neko Health, though it consumes significant time as one of the world's biggest problems. He maintains involvement with Spotify as the largest shareholder, though not on a day-to-day basis. Through Prima Materia, the company he created with Shaq, Ek aims to be the greatest co-founder possible, focusing on paying forward success.

Healthcare costs rank among the top three affordability issues in the US alongside housing and education. The system shows extreme pricing distortions, including $15,000 for ER stitches, $20,000 for drugs costing $30 to manufacture, and $6,000 insurance charges for 8-minute doctor visits despite physicians earning $200,000 annually.

The healthcare system's incentive structure is built around treating infectious diseases, focusing on acute intervention when symptoms become severe. The system needs to shift toward prevention with long-term orientation. Employment-tied healthcare creates misaligned incentives since average employment tenure is short (2-3 years), making insurers reluctant to invest in interventions with 10-20 year payback periods.

Neko Health aims to lower costs so ROI calculations become more feasible for insurers and employers. Rather than requiring million-dollar speculative investments with 20-year payback uncertainty, the model targets tens of thousands of dollars with potentially similar timeframes. The company also focuses on adding more data to the system through multimodal longitudinal datasets to demonstrate efficacy over time.

Neko Health releases annual data findings to demonstrate population-level outcomes across multiple countries. This multi-country approach enables comparisons between healthcare systems, such as findings from the UK versus the US. The company has conducted four clinical trials with two underway and four more planned, partnering deeply with the research community.

Early data from 100,000 scans suggests differences between Stockholm and UK populations, with cardiovascular disease and diabetes appearing as the two largest healthcare spending categories. While population-level health outcome analysis is premature given the small sample size, the company expects to generate meaningful insights as the dataset grows.

The healthcare system contains surprisingly limited data relative to its scale. Increasing available data by 10x or 100x would enable drawing significant conclusions not just for Neko Health but across the entire healthcare system.

Neko Health allows import of Apple Health data, capturing information from wearables including Oura, Whoop, Fitbits, and Apple Watches. The company encourages patients to share wearable data to provide clinicians with 360-degree health perspectives. Integration of wearable metrics with blood work and other diagnostics could transform healthcare delivery.

Unlike traditional diagnostic products requiring years of R&D followed by decade-long sales cycles, Neko Health launched in the US on generation two after only three years. The company plans continuous addition of valuable diagnostics at the current price point, with the goal of delivering the most valuable hour possible on health.

Ek believes society remains in early innings where impact can still be altered. Every great technology presents extreme positives and negatives, requiring conscious steering toward desired outcomes. The industry has failed to adequately highlight positive AI applications, such as Spotify's future capability to soundtrack every moment of life or dermatology applications tracking 950 moles over time.

Ek advocates for open source and open weight models, noting the dramatic cost difference between 13 cents per million token output versus $30. This cost differential enables broad technology access across businesses and individuals, preventing value concentration among few entities. Regulatory conversations may restrict open source expansion.

Technology historically oscillates between open and closed systems, as seen with Windows versus Linux and iOS versus Android. The likely outcome involves both approaches coexisting. Spotify uses both frontier models and fine-tuned proprietary models, achieving greater innovation through this dual approach due to cost, efficiency, and customization advantages.

Compute volume may serve as a better indicator of AI capability than model intelligence alone. Using 100,000 GPUs likely produces more powerful results than running open-source models on home PCs. This compute-based perspective could inform regulatory approaches by focusing on access to significant compute resources rather than model certification.

Spotify supports both open and closed approaches in podcasting. While maintaining proprietary features like platform-specific comments, Spotify also aggregates and distributes podcasts to other platforms through open standards. The company believes both approaches can coexist, though innovating within standards requires broader agreement and may slow development.

Before Spotify, Ek worked at Stardoll, originally Paper Doll Heaven from Finland. Index Ventures invested to scale the company, hiring Ek to rebuild the technical infrastructure. The site had 4-minute page load times due to server issues. Ek rearchitected the website, hired a new technical team, and reduced load times to under one second, causing traffic to explode. After 6-12 months, Ek left to build Spotify.

Keep All-In Podcast in your library

Save the videos and channels worth coming back to, and find them again in one place.