Spotify's Ek Brings $499 Full-Body Scans to US in Healthcare Shift

Daniel Ek: Life After Spotify, Broken Healthcare Incentives, Catching Disease Early & AI's Potential — All-In Podcast

All-In Podcast

US healthcare fails because short-term insurer turnover eliminates incentives to fund long-term disease prevention, requiring vertically integrated clinics to slash upfront diagnostic costs directly for consumers.

AI-assisted summary of Daniel Ek: Life After Spotify, Broken Healthcare Incentives, Catching Disease Early & AI's Potential — original source. Check the original for full context.

Source published:

Key ideas

  • Daniel Ek has transitioned to Executive Chairman of Spotify (>700M users) while actively scaling Neko Health, his preventive healthcare venture opening its first US flagship clinic at 300 Lafayette Street in New York City. (0:00)
  • Neko Health packages 53 on-site blood biomarkers, cardiovascular testing, and a 6,000-image full-body optical scan indexing an average of 950 moles into a 60-minute, $499 consumer clinic visit. (12:35)
  • Ek attributes the US healthcare crisis to structural incentive failure: because American workers change employers every 2–3 years, private insurers have zero financial incentive to reimburse a $500 screen that prevents a $200,000 chronic illness 20 years later. (27:18)
  • On artificial intelligence governance, Ek advocates regulating physical compute scale (such as clusters of 100,000+ GPUs) rather than software weights, warning that the industry has committed a grave error by amplifying apocalyptic risks instead of life-saving applications. (40:00)

Detailed notes

## Background Published on YouTube by the All-In Podcast on September 28, 2026, this episode features Spotify co-founder Daniel Ek reflecting on his career transition to Executive Chairman and his work in preventive healthcare. According to the show notes, the discussion spans the early days of Spotify beating digital piracy, the launch and economics of his preventive healthcare company Neko Health, the potential and regulation of artificial intelligence, and the evolution of open podcasting standards.

## TL;DR Spotify co-founder Daniel Ek reflects on his entrepreneurial journey, detailing how he and Martin Lorentzon underwrote record label executive bonuses with their own money to secure early music licenses before expanding Spotify globally. He explains how he is applying that identical geographic and technological playbook to Neko Health, a preventive diagnostic service offering vertically integrated, $499 multi-modal body scans to catch chronic conditions before symptoms appear. Ek diagnoses the structural failures of US healthcare, where short insurer retention spans eliminate financial incentives for long-term preventive care. Shifting to broader technology trends, Ek advocates for open-weight AI models, criticizes industry fearmongering, proposes regulating physical compute scale rather than software intelligence, and revisits his early technical rescue of Stardoll.

## Atmosphere & style The conversation carries the relaxed, collegiate rhythm of veteran technology operators and longtime acquaintances swapping battle-tested war stories and technical theses. Ek brings a grounded, humble demeanor marked by self-deprecating humor, laughing off the sheer pain tolerance required to build hardware while Jason Calacanis and David Friedberg weave in playful jabs and venture capital commiseration. The dialogue moves fluidly from granular unit economics and macro policy failures to personal reflections, creating an engaging, intellectually rigorous, yet approachable listening experience.

## Bulletin - [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) Daniel Ek has transitioned to Executive Chairman of Spotify (>700M users) while actively scaling Neko Health, his preventive healthcare venture opening its first US flagship clinic at 300 Lafayette Street in New York City. - [12:35](https://www.youtube.com/watch?v=JEUboZzZGM4&t=755s) Neko Health packages 53 on-site blood biomarkers, cardiovascular testing, and a 6,000-image full-body optical scan indexing an average of 950 moles into a 60-minute, $499 consumer clinic visit. - [27:18](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1638s) Ek attributes the US healthcare crisis to structural incentive failure: because American workers change employers every 2–3 years, private insurers have zero financial incentive to reimburse a $500 screen that prevents a $200,000 chronic illness 20 years later. - [40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) On artificial intelligence governance, Ek advocates regulating physical compute scale (such as clusters of 100,000+ GPUs) rather than software weights, warning that the industry has committed a grave error by amplifying apocalyptic risks instead of life-saving applications.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) Spotify co-founder Daniel Ek stepped into the Executive Chairman role after two decades as CEO, having scaled the platform to over 700 million active users and 300 million premium subscribers. - [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) To secure early licensing deals in Sweden, Ek and Martin Lorentzon used personal capital to underwrite record label executives' annual budgets and bonuses, eliminating downside risk for the labels. - [01:14](https://www.youtube.com/watch?v=JEUboZzZGM4&t=74s) Sweden served as the ideal launchpad for Spotify because universal broadband penetration and widespread piracy without an official iTunes Store had collapsed domestic recording revenue by 80%. - [06:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=390s) Neko Health mirrors Spotify's five-year developmental expansion model, moving methodically from quiet R&D and launch in Sweden to the UK, and now entering the United States. - [10:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=600s) Healthcare accounts for 18% of US GDP, with chronic conditions driving the bulk of spending despite being overwhelmingly manageable or preventable when detected before symptoms manifest. - [12:35](https://www.youtube.com/watch?v=JEUboZzZGM4&t=755s) Neko Health provides a vertically integrated full-body assessment for $499, delivering real-time results across 53 blood markers, arterial stiffness, and 360-degree dermatological scans in under an hour. - [14:15](https://www.youtube.com/watch?v=JEUboZzZGM4&t=855s) Empirical survey data across more than 100,000 scans shows that members entering with the poorest objective health markers exhibit the highest rates of lifestyle adherence when shown granular visual proof of risk. - [20:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1200s) Because the average human body has roughly 950 moles, longitudinal computer-vision tracking is essential to detect minute morphological shifts and early-stage melanomas that escape the human eye. - [22:42](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1362s) Ek argues annual full-body preventive checkups should mirror routine dental cleanings, establishing a cultural norm that dramatically bends long-term health spending. - [25:12](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1512s) By developing its own proprietary scanning hardware and software stacks, Neko removes healthcare middleman margins, enabling positive unit economics at consumer price points. - [27:18](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1638s) Modern American healthcare financing was architected for acute infectious disease outbreaks rather than chronic conditions, decoupling long-term preventative health outcomes from insurer profitability. - [35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) Ek criticizes tech industry commentary for indulging in public hubris and hyper-focusing on AI doomsday scenarios instead of championing immediate upsides in diagnostics and quality of life. - [35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) The steep reduction in open-source AI inference costs—from $30 down to $0.13 per million output tokens—is vital to prevent total value capture by a handful of tech incumbents. - [40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) AI safety policy should focus on regulating physical compute clusters like historical Cray supercomputer export controls, rather than policing open-weight software that runs on consumer-grade PCs. - [44:24](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2664s) Spotify maintains a hybrid podcasting approach by feeding open RSS distribution rails while independently developing proprietary platform features to bypass slow standards-committee cycles. - [46:42](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2802s) During his pre-Spotify stint as interim CTO of Stardoll, Ek overhauled backend infrastructure to reduce page load times from four minutes to under one second, securing an early growth round from Sequoia Capital.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) Welcome Daniel Ek and the Origins of Spotify

The *All-In Podcast* hosts Jason Calacanis and David Friedberg introduce Spotify co-founder Daniel Ek, who led the streaming giant across two decades from an ambitious Swedish startup into a global utility boasting well over 700 million active users and more than 300 million premium subscribers. Having transitioned into the role of Executive Chairman on January 1st, Ek joined the hosts to reflect on his journey and introduce his latest venture in preventive healthcare, Neko Health. Jason reminisces about encountering an early demo of Spotify roughly two decades ago at an internet summit, back when consumer subscription models for digital media were virtually nonexistent and the venture seemed completely improbable.

- **Spotify Scale:** Over 700 million monthly active users and more than 300 million paying premium subscribers. - **Leadership Transition:** Ek stepped into the Executive Chairman role on January 1st after leading the company as CEO for twenty years.

---

[01:14](https://www.youtube.com/watch?v=JEUboZzZGM4&t=74s) The Pirate Era and Pitching the Impossible Music Service

Transporting listeners back to Spotify's inception, Ek paints a picture of a radically different digital landscape: an era preceding the iPhone, where YouTube did not exist, and Facebook was strictly confined to college campuses. The global music industry had entered a terrifying free fall fueled by widespread peer-to-peer piracy via Napster, Kazaa, and BitTorrent. In the United States, the Recording Industry Association of America (RIAA) resorted to aggressive legal measures, suing individual consumers for illegally downloading MP3s. Watching the chaos unfold from Sweden, Ek recognized that suing fans was futile because "there's no way to put the genie back in the bottle."

While brainstorming startup concepts with his co-founder, Martin Lorentzon, Lorentzon asked Ek what he would pursue if he could do anything in the world. Ek admitted his real passion was music, but immediately dismissed it as "a really stupid idea" due to the legal minefield of licensing, intellectual property clearances, and negotiating with notoriously rigid record labels. Lorentzon relentlessly countered with a series of "what if" questions until Ek ran out of objections, landing on a foundational product vision: an experience so seamless and responsive that "it would feel like you had all the world's music at your fingertips."

While industry insiders laughed off their initial ambition to launch a global music service from day one, Sweden proved to be an unexpected testing ground. Because the country had rolled out high-speed broadband in the early 2000s and Apple's iTunes Store was not yet available locally, piracy ran rampant with zero viable legal digital alternatives. As a consequence, the domestic music recording business had collapsed by an estimated 80%, leaving local labels desperate enough to listen to a radical proposal.

---

[05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) The Label Deal: Guaranteeing Bonuses and the Five-Year Path to the US

To break through the label executives' paralysis and fear, Ek and Lorentzon devised a high-stakes, unconventional bargaining chip by putting their own personal capital from previous startup exits on the line. They approached the labels with a downside-protection guarantee: Spotify would financially underwrite the executives' budgets and bonuses for the upcoming year out of pocket. Their pitch was simple: "If this really is as bad as you think it is, then you just shut it down after a year, but you still made the bonus. And if this actually works, you win as well. So you win no matter what."

Drawn in by a proposal that eliminated personal financial risk, the labels eventually signed on after years of grueling negotiations. Spotify launched in late 2008 in Sweden, rapidly achieving explosive product-market fit before expanding into the United Kingdom, and ultimately arriving in the United States in 2011—a five-year journey from concept to American launch.

```mermaid flowchart LR accTitle: The Spotify and Neko Health Expansion Playbook accDescr: A three-stage geographic expansion model moving from an initial market in Sweden to the UK and then the US.

subgraph MarketPhases [" Geographic Expansion Model "] direction LR A["Sweden<br>R&D & Initial Validation"] --> B["United Kingdom<br>Secondary Expansion"] B --> C["United States<br>Scaled Global Market"] end ```

- **Incentive Alignment:** Ek and Lorentzon guaranteed music executives' annual budgets and performance bonuses using their own funds to de-risk the licensing experiment. - **Rollout Timeline:** Founded in the mid-2000s, Spotify launched publicly in Sweden in late 2008, expanded to the UK shortly thereafter, and entered the US market in 2011.

---

[06:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=390s) Applying the Spotify Playbook to Neko Health and Why Ek Prefers Building to Investing

When David Friedberg notes the sheer grit required to spend five years in the laboratory before even entering the United States, Ek reveals that he is running the exact same geographic playbook with his preventive healthcare company, Neko Health. Ek co-founded Neko eight years ago in 2018, spending five years engineering the proprietary scanning technology before opening in Sweden in 2023, expanding to the UK, and now preparing for a US debut. Ek laughs that whether it is sheer entrepreneurial drive or simply an abnormally "high pain tolerance," he feels compelled to tackle massive systemic dysfunctions. Although most technological trends are progressing "up and to the right," he views healthcare as a rare, broken sector where costs escalate while outcomes languish.

Ek's focus on healthcare is not a sudden whim; his team recently reminded him of a *Financial Times* "Lunch with the FT" interview from around 2012 or 2013 where he spoke passionately about reinventing healthcare right after Spotify's US rollout. He explains that he was not an active angel investor throughout most of his career because virtually all his net worth remained tied up in Spotify until its direct listing in 2018. When he finally experimented with angel investing and board seats, he quickly discovered he despised the sidelines. Calacanis, Friedberg, and Ek commiserate on the agony of watching portfolio companies get mismanaged from a boardroom seat: if a founder ignores board advice, you tear your hair out; if a founder blindly follows your instructions, they lack genuine leadership. Consequently, Ek resolved that his time is best spent building companies directly, adhering to the angel philosophy of backing extraordinary founder-CEOs and leaving them completely alone.

- **Neko Health Timeline:** Founded in 2018; spent five years in quiet hardware and software development; launched clinics in Sweden in 2023, followed by the UK, with the US next. - **Liquidity & Focus:** Ek had no personal liquidity until Spotify's 2018 public offering, keeping him entirely focused on operating Spotify for 12 to 13 years before writing early-stage checks. - **Boardroom Dilemma:** The hosts and Ek agree that directing CEOs from board seats fails because strong companies require autonomous founder leadership rather than outsourced managerial guidance.

[10:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=600s) Origins of Neko Health: From Chronic Disease to the Data Flywheel

The conversation turns to Daniel Ek's motivation for stepping directly into the healthcare arena rather than advising from the sidelines, underscoring that founders with true agency must build solutions themselves. Ek frames the scale of the crisis by pointing to macroeconomic fundamentals: the United States spends 18% of its GDP on healthcare, making it the single largest line item in the entire budget. Looking at heart disease in isolation, annual spending reaches hundreds of billions of dollars in the US alone—an amount eclipsing the annual revenues of Fortune 10 corporations for a single disease category. While researching everything from oncology to pharmaceutical efficiency, Ek concluded that the dominant burden stems from chronic disease. Crucially, chronic conditions are overwhelmingly preventable if caught in their earliest stages, where intervention costs are negligible and patient suffering is virtually eliminated.

The central barrier to early discovery is a profound lack of data. While the entire healthcare establishment universally agrees that medicine must transition from a reactive model to a preventative one, industry players fundamentally diverge on execution. Ek and his co-founder, Hjalmar Nilsonne, anchored their philosophy on building a predictive, multi-modal data flywheel. To anticipate adverse health events and execute early course corrections, medicine requires longitudinal tracking across several distinct bodily systems. Ek identified two technological convergences that made this possible: the smartphone revolution—which dramatically drove down the cost of advanced sensors and miniaturized electronics—and modern artificial intelligence. Drawing an explicit parallel to his core work, Ek notes that the AI recommendation engines at Spotify rely on massive data ingestion to predict what song a listener wants next; applying that identical machine learning thesis across multi-modal, longitudinal health metrics provides the foundation for accurate predictive medicine.

```mermaid flowchart TD accTitle: Neko Health Predictive Flywheel accDescr: Flowchart illustrating how multi-modal sensors and AI drive preventative healthcare over time.

subgraph INGESTION ["Data Capture"] SENSORS["Low-Cost Miniaturized Sensors<br/>(Smartphone Hardware Tailwinds)"] MODELS["Multi-Modal Inputs<br/>(Dermatology, Blood, Cardio)"] end

subgraph ENGINE ["Intelligence Layer"] LONGITUDINAL["Longitudinal Data Tracking<br/>(Measurements Over Time)"] AI["AI & Pattern Recognition<br/>(Spotify-Style Predictive Models)"] end

subgraph OUTCOME ["Preventative Intervention"] PREDICT["Early Anomaly Detection"] PREVENT["Low-Cost Lifestyle & Clinical Remedies<br/>(Preventing Chronic Disease Progression)"] end

SENSORS --> MODELS MODELS --> LONGITUDINAL LONGITUDINAL --> AI AI --> PREDICT PREDICT --> PREVENT PREVENT -.->|"Refines Baselines"| LONGITUDINAL ```

- **US Healthcare Burden:** Healthcare consumption absorbs roughly 18% of US GDP, standing as the nation's largest single budgetary expenditure. - **Cardiovascular Costs:** US spending on heart disease alone exceeds hundreds of billions of dollars annually, surpassing the revenue scale of top Fortune 10 conglomerates. - **The Preventative Consensus:** Industry consensus agrees on moving from reactive treatment to prevention, but legacy models lack the continuous data pipelines required to anticipate disease onset.

---

[12:35](https://www.youtube.com/watch?v=JEUboZzZGM4&t=755s) The $499 Experience: Reinventing Healthcare from the Ground Up

When prompted on whether observational screening can alter disease trajectories beyond baseline lifestyle management, Ek walks through the actual physical product experience of Neko Health Gen 2. Built as a vertically integrated, ground-up healthcare service priced at $499, Neko's stated objective is to deliver "the most valuable hour you can invest in your health." Rather than navigating fragmented referrals, separate laboratories, and weeks of anxiety waiting for disparate charts, the patient undergoes an all-inclusive, one-hour continuous clinical scan.

The appointment begins immediately with an on-site blood draw analyzing 53 distinct blood markers. Next, the patient steps into Neko's proprietary 360-degree skin-scanning rig, which leverages an advanced high-resolution multi-camera array to capture upwards of 6,000 microscopic images of the entire body. The system computationally maps and indexes every single mole, lesion, rash, and instance of redness. The visit subsequently moves through cardiovascular and circulation diagnostics, along with foundational baseline markers like grip strength, which hold proven scientific validity as proxies for longevity and biological age. Crucially, the visit concludes with uninterrupted, face-to-face consultation time with a licensed clinician who reviews the synthesized diagnostic package—including the real-time blood test results processed during the visit—and formulates actionable next steps. This model is launching initially in New York before rolling out across major US metropolitan markets.

- **Unit Economics & Time:** A complete, end-to-end full-body evaluation priced at $499, completed within a single 60-minute visit. - **Diagnostic Breadth:** 53 integrated blood biomarkers evaluated on-site alongside cardiovascular, arterial stiffness, and physical metrics (such as grip strength). - **Computer Vision Rig:** Over 6,000 ultra-high-resolution images taken across the entire surface of the body to catalogue dermatological anomalies. - **Real-Time Integration:** Clinician review takes place immediately at the conclusion of the scan, reviewing blood and sensor data simultaneously with the member.

---

[14:15](https://www.youtube.com/watch?v=JEUboZzZGM4&t=855s) Measurable Outcomes and the Power of Visual Health Data

Addressing comparisons to prior venture-backed concierge clinic attempts such as Forward and One Medical, Ek emphasizes Neko's extensive real-world clinical validation. Founded in 2018, Neko has refined its technology over eight years, conducting more than 100,000 scans and publishing longitudinal health survey data annually. Now in its third consecutive year of peer-reviewed data release, the results demonstrate that Neko regularly uncovers previously undiagnosed, serious underlying medical conditions among seemingly asymptomatic members.

While healthy individuals use the service to optimize sleep, diet, and stress management, Ek highlights an unexpected behavioral breakthrough: members presenting with the worst baseline health markers demonstrate the most dramatic, sustained improvements. Seeing high-fidelity, visual data of their own arterial health, systemic inflammation, or skin risk laid bare—paired with a direct clinician roadmap—shatters patient inertia and catalyzes urgent lifestyle modifications. To illustrate the scale of hidden dermatological complexity, Ek shares a startling statistic: the average person scanned at Neko possesses 950 individual moles, a volume that makes self-examination and manual spot-checks practically impossible without automated computer vision.

- **Scale and Longevity:** Conceived in 2018; accumulated eight years of operational deployment and over 100,000 comprehensive scans completed. - **Empirical Publications:** Annual data surveys published three years running, tracking real-world health outcomes and undiagnosed condition detection. - **Behavioral Shift:** Patients entering with the poorest objective health metrics show the highest rate of lifestyle adherence and metric improvement due to granular visual clarity. - **Dermatology Demographics:** Scans reveal that the average human body carries approximately 950 moles, rendering unassisted self-screening ineffective.

[20:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1200s) AI Mole Mapping and Longitudinal Tracking

Daniel Ek explains the stark limitations of traditional dermatological exams when compared to algorithmic, full-body imaging. Human skin varies widely from person to person—some individuals have roughly 500 moles, while others have thousands—making it virtually impossible for the naked human eye or a physician's memory to track subtle abnormalities over time. In a typical clinic visit, a doctor visually scans for obvious irregularities, zooms in on a suspicious spot with a dermatoscope, and might capture cross-frequency images to analyze size and texture. Neko Health automates and scales this diagnostic path by pairing broad-spectrum optical scanning with a multi-tiered clinical review.

```mermaid flowchart TD accTitle: Neko Health Scanning and Review Pipeline accDescr: Step-by-step screening pipeline showing data capture, AI analysis, clinical escalation, and annual longitudinal tracking.

subgraph Capture["1. Scanning & Analysis"] A["Full-Body Multi-Frequency Scan"] --> B["AI Detection Engine"] end

subgraph Triage["2. Clinical Review"] B -->|"Flag Risk Factors"| C["On-Site Human Clinician"] C -->|"Suspicious Findings"| D["Staff Dermatologists"] end

subgraph History["3. Longitudinal Monitoring"] C --> E["Year-over-Year Digital Archive"] E -->|"Compare Growth & Texture"| B end ```

The true power of this system emerges through longitudinal data. When members return for subsequent scans, Neko’s software compares high-resolution, indexed records of every single mole against prior baselines. Even elite dermatologists cannot recall the minute dimensions or surface textures of a patient's mole from twelve months earlier. By systematically flagging abnormal expansion or morphological shifts year over year, the system catches early-stage melanomas and skin cancers before they become life-threatening. When asked whether he runs the company day-to-day, Ek clarifies his role with self-deprecating humor: much like Spotify—where leadership is shared—Neko is spearheaded by his co-founder and CEO, Hjalmar Nilsonne, whom Ek credits as the operational mastermind while joking that he merely sits on podcasts taking the credit.

[22:42](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1362s) The $499 Annual Checkup and Preventive Health Economics

The discussion shifts to the economics of preventive screening, comparing Neko Health's $499 comprehensive full-body scan against consumer lab services like Function Health and Superpower (which charge $350 to $400 exclusively for deep blood panels). With US federal entitlement spending dominated by Social Security and healthcare—closely followed by escalating interest payments on the national debt—the hosts question whether shifting a fraction of that expenditure toward early detection could bend the long-term cost curve of macro healthcare spending.

While Ek notes it is still too early to present conclusive epidemiological datasets on national budgetary impact, he emphasizes that annual preventive checkups ought to be cultural baseline habits. Drawing a direct parallel to dental hygiene, Ek points out the societal paradox where people routinely visit a dentist twice a year to maintain their teeth, yet rarely schedule an annual preventive evaluation for the rest of their body. Neko is designed around maximizing patient compliance by creating "the most valuable hour you can invest in your health"—a brutally efficient experience that can take significantly less than sixty minutes from arrival to review if the patient has no outstanding questions.

- **Screening Price Comparison:** Neko Health provides full-body imaging and cardiovascular mapping for $499, competing in value with diagnostic blood services like Function Health and Superpower ($350–$400). - **Time Commitment:** Optimized for brutal efficiency, completing data collection, imaging, and clinician review in under an hour. - **Societal Disconnect:** Annual checkups remain rare among healthy working adults despite widespread adoption of routine annual dental exams.

[25:12](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1512s) Vertical Integration and Daniel Ek's Portfolio

Addressing skepticism around whether a $499 price point represents an unsustainable loss leader designed to upsell patients, Ek explains that Neko Health operates on positive unit economics. Rather than stitching together disparate third-party medical devices, proprietary software licenses, and outsourced laboratories, Neko designed and built its own integrated scanning hardware and software stacks from the ground up. This vertical integration strips out traditional healthcare middleman margins, enabling existing clinics to run profitably at retail pricing while generating the capital required to fund international clinic expansion.

When asked about his daily allocation of time, Ek outlines his current balance across operational and investment commitments. While he no longer manages Spotify on an everyday operational level, he continues to serve actively as Executive Chairman, affectionately referring to the streaming giant as his "baby." Beyond Spotify and Neko Health, Ek allocates significant energy to Prima Materia—the European investment partnership he founded alongside Shakil Khan ("Shak")—which aims to act as the ultimate co-founder for moonshot founders tackling hard scientific and societal challenges.

[27:18](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1638s) Broken Incentives in the US Healthcare System

The hosts confront Ek with the stark realities of the American healthcare affordability crisis, contrasting it with housing and education. While market mechanisms can theoretically resolve housing supply and educational costs, the US medical system routinely produces absurd price distortions: an emergency room visit for simple stitches billing at $15,000, off-patent drugs costing $30 to manufacture being priced at $20,000, or an eight-minute physician consultation generating a $6,000 insurance charge despite the doctor earning an average salary of $200,000.

Invoking Charlie Munger’s dictum—"show me the outcome, I'll show you the incentive"—Ek diagnoses the root structural failure of modern medicine:

- **Infectious Disease Legacy:** Modern medical delivery and billing frameworks were designed decades ago to manage acute, symptomatic infectious outbreaks rather than silent, slow-moving chronic illnesses. - **Short-Term Payer Horizon:** Because American health insurance is tethered directly to employment, and the average worker changes jobs every two to three years, health plans change just as frequently. - **Zero Preventative ROI for Insurers:** A private insurer has no financial incentive to reimburse a $500 comprehensive screen that prevents a $200,000 cardiovascular event fifteen to twenty years later, because by the time the disease manifests, the patient will be covered by an entirely different insurer or Medicare.

[30:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1800s) Lowering Costs and Multimodal Longitudinal Data in Healthcare

Addressing the core financial challenge of preventive medicine, Daniel Ek highlights the struggle of calculating ROI on healthcare investments that might only pay off 10, 15, or 20 years down the road. Traditionally, neither employers nor private health insurers want to risk millions of dollars on speculative prevention with minimal upfront data. Neko Health's strategy tackles this by radically driving down hardware and operational costs. If preventive scanning costs only tens of thousands of dollars across a population rather than millions, payer propensity to adopt it skyrockets, even under extended payback horizons.

Alongside affordability, Neko Health is building large, multimodal longitudinal data sets to prove diagnostic efficacy over time. Ek notes that Neko is committed to publishing its clinical findings every year. Operating across both the UK and Sweden makes Neko one of the rare early-stage health tech companies capturing multi-country comparative data from day one. When Jason Calacanis jokingly assumes that the Stockholm cohort is far fitter while the UK struggles with cardiovascular disease and diabetes—prompting David Friedberg to note that cardiovascular disease and diabetes represent the two largest healthcare cost drivers—Ek explains that it is still too early to declare population-level conclusions on their base of 100,000 scans. However, Neko has already completed four clinical trials, has two active, and plans four more with the academic and medical research community.

```mermaid flowchart TD accTitle: Neko Health Longitudinal Data Pipeline accDescr: Diagram illustrating how multimodal inputs from scans, wearables, and clinical trials feed into longitudinal datasets to lower preventive care costs.

subgraph DataCollection["Data Ingestion"] A["Neko Full-Body Scan"] B["Apple Health & Wearables<br/>(Oura, Whoop, Fitbit)"] C["Clinical Trials & Labs<br/>(10 Completed/Planned)"] end

subgraph Processing["Aggregated Insights"] D["Multi-Country Longitudinal Dataset<br/>(Sweden, UK, US)"] E["Doctor Consultation<br/>(360-degree Patient View)"] end

subgraph ValueOutput["System-Wide Impact"] F["Annual Public Data Release"] G["Reduced Diagnostic Costs & High-ROI Prevention"] end

A --> D B --> D C --> D D --> E D --> F E --> G F --> G ```

- Daniel emphasizes that tech outsiders are often shocked by how small healthcare data sets actually are; multiplying that data by 10x or 100x could transform medical understanding across the entire sector. - When asked if Neko plans to build its own hardware wearable or partner with existing trackers like Oura, Whoop, Fitbit, or Apple Watch, Daniel notes Neko already ingests Apple Health data so clinicians can review biometric histories alongside scans to provide a 360-degree view.

[35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) Rapid Iteration in Diagnostics and AI's Blind Spots

Traditional diagnostic products follow an archaic lifecycle: teams spend years in R&D and then sell the exact same static hardware device for a decade. In contrast, Neko Health operates like a consumer software and tech company, launching Gen 2 infrastructure within three years as it expands into the United States. Ek aims to pack progressively richer diagnostic data into the same price point, pursuing the company's core mission: delivering "the most valuable hour you can spend on your health."

The conversation shifts to broader artificial intelligence debates and whether society must artificially pace frontier models. Ek observes that over recent years, public commentary has suffered from widespread hubris:

> "We're walking into COVID and every one of us became a virologist. We were all of a sudden supply chain experts after the first oil tanker in the Hormuz Strait got stuck somewhere. And then with the Ukraine war breaking out, we were all military experts and military strategists."

Jason Calacanis and David Sacks tease that Daniel is describing the *All-In Podcast* itself—joking they should rename it the "All-In Laboratory Pod" whose opinions need Spotify podcast regulation. Daniel stresses that loud confidence often precedes being completely wrong. While he maintains no dogmatic stance on whether to artificially slow frontier AI, he laments that the tech industry has done a "terrible disservice" by hyper-focusing on apocalyptic risks instead of marketing the enormous positive upsides: from Neko tracking 950 skin moles over time to keep patients cancer-free, to Spotify leveraging AI so listeners can soundtrack every moment of their lives and "feel more" emotionally.

- Daniel strongly endorses open-source AI and open-weight models, arguing that dropping inference costs from $30 down to $0.13 per million output tokens democratizes power across thousands of businesses, preventing the extreme concentration of value and wealth in a handful of tech incumbents.

[40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) Open vs. Closed Models and Regulating AI Compute

Addressing the ongoing debate over open weights versus proprietary frontier systems, Daniel Ek observes that technology has historically swung between open and closed paradigms. Just as personal computing settled into Windows versus Linux and mobile ecosystems into iOS versus Android, Ek believes artificial intelligence will inevitably support both models. At Spotify, engineering teams rely heavily on closed frontier models while simultaneously deploying internally fine-tuned, open-weight models. Ek explains that this dual approach accelerates practical innovation, granting them specialized performance, cost efficiencies, and bespoke tuning options that closed commercial APIs cannot deliver.

Turning to AI safety and governance, Ek introduces a dimension he believes has been largely neglected in policy discussions: regulating total compute rather than software intelligence in the abstract. Evaluating real-world risk requires looking at the raw horsepower behind an operation. An adversary running an open model locally on a consumer PC faces steep technological barriers against high-grade enterprise defenses, whereas an operation marshaling 100,000 GPUs possesses systemic offensive capabilities.

```mermaid flowchart TD accTitle: Compute-Based AI Risk Assessment accDescr: Diagram contrasting consumer-level compute risk against enterprise-scale GPU clusters.

subgraph Scale["Operational Compute Scale"] PC["Local Consumer Hardware<br/>(Single PC / Consumer GPU)"] Cluster["Massive Enterprise Compute<br/>(100,000+ GPUs)"] end

subgraph Impact["Risk and Defensive Posture"] LowRisk["Limited threat capability<br/>Blocked by frontier cyber defenses"] HighRisk["Severe systemic threat profile<br/>Requires strict hardware certification"] end

PC --> LowRisk Cluster --> HighRisk ```

The hosts compare compute-based oversight to historical regulatory regimes for Cray supercomputers, which were treated as national security assets and restricted under strict licensing because of their potential to design nuclear weapons and crack cryptographic codes. They contrast this structural hardware control with clumsy legislative proposals—such as early California state bills—that attempted to classify dangerous software by static thresholds of training teraflops, data-center power wattage, or network cabling. Assuming parity between underlying model architectures, sheer installed compute capacity becomes the primary defensive moat.

- Historical operating system and mobile platform splits show that open and closed architectures consistently co-exist in equilibrium. - Spotify leverages frontier proprietary models for general intelligence and bespoke fine-tuned models for domain-specific tasks and cost efficiency. - Ek argues compute consumption serves as a practical, measurable proxy for catastrophic cyber risk compared to policing open-source model weights alone.

---

[44:24](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2664s) Podcasting Open Standards and Spotify's Platform Strategy

Jason Calacanis challenges Ek on Spotify's relationship with podcasting, an industry built on decentralized, open RSS standards that has increasingly bumped up against Spotify's walled-garden features. Ek explains that Spotify deliberately maintains a hybrid strategy. While Spotify functions as an end-user destination with native, platform-specific capabilities—such as dedicated interactive comment sections—it simultaneously operates creator infrastructure that packages, aggregates, and syndicates podcasts across open third-party platforms. Ek notes that developing features strictly within open committee standards requires universal ecosystem alignment, which dramatically slows development velocity. To bypass that stagnation, Spotify develops proprietary platform enhancements while continuing to feed the open standard.

```mermaid flowchart LR accTitle: Spotify Dual Podcasting Architecture accDescr: Flowchart illustrating Spotify's support for open RSS syndication alongside proprietary platform features.

Creator["Podcast Creator"] --> SpotifyCore["Spotify Creator Tools"]

subgraph Open["Open Ecosystem"] SpotifyCore -->|"Open RSS Standard"| ExtPlatforms["Apple Podcasts, Overcast, etc."] end

subgraph Proprietary["Native Ecosystem"] SpotifyCore -->|"Proprietary Platform"| AppDestination["Spotify Consumer App<br/>(Interactive comments, video, analytics)"] end ```

Representing longtime independent podcasters, Calacanis pitches two specific open-standard features for Spotify to adopt: 1. **Live status signaling:** Parsing RSS tags that alert podcast clients when an episode is actively streaming live. 2. **Value-for-value donation rails:** Native support for podcast monetization and listener tipping via protocols like PayPal or lightning micropayments.

Ek laughs that while product roadmaps are no longer his unilateral decision as executive chairman, he will personally carry the feature requests to Spotify's product and engineering teams.

---

[46:42](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2802s) The Stardoll Origin Story: From Four-Minute Page Loads to Sequoia

Calacanis shifts the discussion to the early Swedish startup scene, asking Ek about his pre-Spotify tenure at Stardoll—one of the first Stockholm tech companies to attract premier American venture capital. Ek recalls joining the company just before conceiving Spotify. The venture began life as "Paperdoll Heaven," a simple web portal built in Turku, Finland, by an older woman and her son. Swedish serial entrepreneur Mattias Miksche—an established figure from the late-1990s dot-com era—partnered with Danny Rimer of Index Ventures to acquire a majority stake, rebrand the site as Stardoll, and scale it into an interactive digital fashion and paper-doll platform.

Though Ek had already resolved to start Spotify, Rimer persuasively lobbied him to step in as interim CTO on a 6-to-12-month tour of duty. Traveling to Finland to inspect the infrastructure, Ek discovered a site buckling under immense organic traffic: - Average page load times hovered around a staggering four minutes per page as backend servers collapsed. - Ek wiped and re-architected the technical platform from scratch, recruiting a dedicated software engineering team. - The re-engineering reduced page load times from four minutes to under one second, causing platform traffic and user engagement to surge.

The performance turnaround attracted a major growth investment round from Sequoia Capital. With the infrastructure stabilized and institutional funding secured, Ek fulfilled his commitment, handed off operations, and departed to build Spotify full-time.

---

[49:50](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2990s) Neko Health's US Rollout and Wrap-Up

In closing, Ek provides an update on Neko Health, his preventive healthcare venture specializing in full-body sensor scans designed to detect disease before symptoms manifest. The company has formally established its first American flagship clinic in New York City: - **Location:** 300 Lafayette Street (at the intersection of Houston Street) in Manhattan. - **Rollout Timeline:** Expanding clinic capacity across New York City before launching upcoming hubs in Miami, Florida, and Washington, D.C. - **Long-term Roadmap:** Rolling out diagnostic centers across major metropolitan markets nationwide throughout the next 12 to 24 months.

The hosts reflect on the nostalgia of the Houston Street neighborhood, remind listeners to join the diagnostic waitlist at NekoHealth.com, and thank Ek for his time before concluding the episode.

## Notable quotes > [01:14](https://www.youtube.com/watch?v=JEUboZzZGM4&t=74s) there's no way to put the genie back in the bottle. — Daniel Ek

> [01:14](https://www.youtube.com/watch?v=JEUboZzZGM4&t=74s) a really stupid idea — Daniel Ek

> [01:14](https://www.youtube.com/watch?v=JEUboZzZGM4&t=74s) it would feel like you had all the world's music at your fingertips. — Daniel Ek

> [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) If this really is as bad as you think it is, then you just shut it down after a year, but you still made the bonus. And if this actually works, you win as well. So you win no matter what. — Daniel Ek

> [12:35](https://www.youtube.com/watch?v=JEUboZzZGM4&t=755s) the most valuable hour you can invest in your health. — Daniel Ek

> [27:18](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1638s) show me the outcome, I'll show you the incentive — Charlie Munger

> [35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) We're walking into COVID and every one of us became a virologist. We were all of a sudden supply chain experts after the first oil tanker in the Hormuz Strait got stuck somewhere. And then with the Ukraine war breaking out, we were all military experts and military strategists. — Daniel Ek

> [35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) terrible disservice — Daniel Ek

> [35:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2100s) feel more — Daniel Ek

## Mentioned Daniel Ek, Martin Lorentzon, Hjalmar Nilsonne, Jason Calacanis, David Friedberg, David Sacks, Mattias Miksche, Danny Rimer, Shakil Khan, Charlie Munger, Spotify, Neko Health, Prima Materia, Stardoll, Paperdoll Heaven, Index Ventures, Sequoia Capital, All-In Podcast, Financial Times, Napster, Kazaa, BitTorrent, RIAA, Apple iTunes Store, Apple Health, Apple Watch, Oura, Whoop, Fitbit, Forward, One Medical, Function Health, Superpower, PayPal, Cray supercomputers

## Video description (0:00) Daniel Ek joins the Besties! (1:14) Starting Spotify: Beating piracy and betting his own money on the record labels (6:10) Neko Health: The Spotify playbook for healthcare, the $499 body scan, and why US healthcare waits until you're sick (36:04) AI: Tech's failure to sell the upside, open vs. closed models, and regulating compute (44:24) Back to Spotify: Podcasting's open standards and the Stardoll origin story

Thanks to our partners for making this possible!

As wealth grows, complexity compounds. Between investments, tax strategy and estate planning, coordinating it all can feel like a full-time job. Creative Planning can help with the heavy lifting. https://ad.doubleclick.net/ddm/trackclk/N2598215.3565131ALLIN/B36182939.452362736;dc_trk_aid=646454825;dc_trk_cid=260464105;dc_lat=;dc_rdid=;tag_for_child_directed_treatment=;tfua=;ltd=;dc_tdv=1

Starting a business? Northwest Registered Agent gives you everything you need to build a complete Business Identity including free tools and built-in privacy. Get more at https://www.northwestregisteredagent.com/ALLINFREE

Follow the besties: https://x.com/chamath https://x.com/Jason https://x.com/DavidSacks https://x.com/friedberg

Follow on X: https://x.com/theallinpod

Follow on Instagram: https://www.instagram.com/theallinpod

Follow on TikTok: https://www.tiktok.com/@allin

Follow on LinkedIn: https://www.linkedin.com/company/allinpod

Intro Music Credit: https://rb.gy/tppkzl https://x.com/yung_spielburg

Intro Video Credit: https://x.com/TheZachEffect

#allin #tech #news

[Source](https://www.youtube.com/watch?v=JEUboZzZGM4)

中文简报

Spotify创始人Ek跨界进军医疗，将499美元全身扫描引入美国 美国医疗体系的失灵在于短期保险更迭消除了资助长期疾病预防的动力，因而需要垂直整合的诊所直接为消费者大幅削减前期诊断成本。 Spotify联合创始人Daniel Ek回顾了他的创业历程，详细讲述了他与Martin Lorentzon如何在早期自掏腰包为唱片公司高管的奖金兜底以获得音乐版权许可，随后将Spotify推向全球。他解释了自己如何将完全相同的地域和技术策略应用于Neko Health——一项提供垂直整合、定价499美元多模态全身扫描的预防性诊断服务，旨在症状出现前发现慢性疾病。Ek剖析了美国医疗体系的结构性缺陷，即保险机构较短的参保周期抹杀了开展长期预防性护理的财务动力。谈及更广泛的技术趋势，Ek主张支持开放权重AI模型，批评了业内的恐慌煽动，提议监管物理算力规模而非软件智能，并回顾了他早年对Stardoll进行技术援助的经历。 Daniel Ek已转任拥有超过7亿用户的Spotify执行董事长，同时正积极拓展其预防性医疗初创公司Neko Health的规模，该公司在美国的首家旗舰诊所将开设于纽约市拉斐特街300号。 (0:00) Neko Health将53项现场血液生物标志物检测、心血管检查以及包含6,000张图像、平均索引950颗痣的全身光学扫描，整合为单次60分钟、售价499美元的面向消费者的门诊服务。 (12:35) Ek将美国医疗危机归咎于结构性激励机制的缺失：由于美国员工每2至3年就会更换雇主，商业保险机构缺乏任何财务动力去报销一项500美元的筛查，以预防20年后可能产生的20万美元慢性病治疗费用。 (27:18) 在人工智能治理方面，Ek主张监管物理算力规模（例如10万张以上GPU的集群）而非软件权重，并警告称该行业放大末日级风险却忽视拯救生命的应用，犯下了严重错误。 (40:00)

Reading summary

Spotify co-founder Daniel Ek argues that the systemic burden of chronic disease persists due to misaligned US healthcare incentives and outdated delivery models, necessitating vertically integrated, AI-assisted preventative screening. Drawing on Spotify's expansion playbook, Ek contends that lowering diagnostic costs and tracking longitudinal multimodal health data offer a path toward bending long-term medical cost curves. Applying Spotify's Expansion Playbook to Healthcare (6:30) Having transitioned to Executive Chairman after twenty years leading Spotify to over 700 million active users and 300 million paying subscribers, Daniel Ek is directing his operational focus toward preventative healthcare with Neko Health. When establishing Spotify, Ek and co-founder Martin Lorentzon resolved music industry licensing deadlocks by personally underwriting record label executives' budgets and bonuses to eliminate their downside risk. Ek is now applying the same phased geographic expansion model—validating in Sweden, expanding to the United Kingdom, and scaling into the United States—to healthcare diagnostics. Ek explains that he prefers direct operational company building over passive angel investing or board advisory seats, which he found frustrating due to governance constraints. Viewing healthcare as an anomalous sector where technology has failed to bend the cost curve, Ek committed eight years of hardware and software engineering alongside co-founder and CEO Hjalmar Nilsonne before commercializing Neko Health. Incentive Failures in US Chronic Care (27:18) Ek diagnoses the US healthcare crisis—which consumes 18% of GDP, with heart disease alone exceeding hundreds of billions in annual expenditures—as a structural incentive failure. Modern healthcare infrastructure and fee-for-service billing systems were designed decades ago to treat acute, symptomatic infectious outbreaks rather than silent, slow-moving chronic conditions that develop over decades. This structural mismatch is compounded by employer-tied insurance. Because American workers switch employers every two to three years, health plans turn over at a similar rate. As a result, private insurers have zero financial incentive to reimburse a $500 screening that prevents a $200,000 cardiovascular event fifteen to twenty years later, because the financial benefit accrues to a competitor or Medicare. Ek argues that shifting medicine from reactive treatment to prevention requires radically driving down screening costs so that preventative interventions yield positive returns within compressed payer horizons. Vertically Integrated Diagnostics and Longitudinal Flywheels (12:35) Neko Health addresses diagnostic fragmentation by providing an all-inclusive, one-hour clinical appointment priced at $499. The visit integrates an on-site blood draw evaluating 53 biomarkers, non-invasive cardiovascular and arterial stiffness checks, baseline physical metrics like grip strength, and an immediate clinician consultation. A central component is a 360-degree optical rig capturing over 6,000 microscopic images to map every lesion, rash, and mole across the body, where the average patient presents with 950 individual moles. Rather than stitching together third-party medical equipment with marked-up software, Neko vertically integrated its proprietary hardware and software stacks, enabling clinics to operate profitably at retail prices. Across more than 100,000 completed scans and three years of published health data, Ek notes that presenting patients with granular, visual representations of their internal health breaks behavioral inertia, yielding the most significant lifestyle adherence among those with the poorest baseline metrics. AI Infrastructure, Model Balance, and Compute Oversight (40:00) Ek draws a parallel between Neko's predictive health models and Spotify's recommendation algorithms, noting that both depend on massive multimodal datasets to identify subtle patterns. Looking at artificial intelligence broadly, Ek projects that the market will stabilize into a dual equilibrium of proprietary frontier platforms and open-weight models, akin to historical splits between Windows and Linux or iOS and Android. Spotify itself pairs closed commercial APIs for general reasoning with fine-tuned open-source models to reduce inference expenses. On the regulatory front, Ek cautions against catastrophic panic and superficial software regulations. He argues that meaningful AI safety policies must focus on physical compute infrastructure rather than policing open software weights. In Ek's assessment, malicious actors deploying open models on consumer hardware face steep defensive hurdles, whereas systemic security risks concentrate in large-scale deployments marshaling clusters of 100,000 or more GPUs. While Neko Health has completed 100,000 scans across Sweden and the UK and opened its first US clinic in New York City, Ek acknowledges that broad epidemiological datasets proving macro-level budgetary savings across entire national populations remain years away. The ultimate test of the model rests on whether low-cost, consumer-paid full-body screening can overcome the entrenched institutional inertia of employer-based healthcare reimbursement.

精炼阅读摘要

Spotify 联合创始人 Daniel Ek 认为，美国医疗体系因保险绑定雇主与短期利益导向而陷入慢性病治理的结构性失效；他正尝试将流媒体领域的算法预测和硬件垂直整合经验引入预防医疗，通过低成本、多模态的全身早筛与数据追踪来重塑疾病干预模式。 创业路径：从流媒体突破到预防医疗构建 (6:30) 回顾 Spotify 的创立过程，Daniel Ek 指出其核心策略在于打破利益僵局与循序渐进的市场验证。在 2000 年代数字音乐盗版泛滥之际，Ek 与联合创始人 Martin Lorentzon 动用以往创业退出的个人资金，直接为唱片公司高管担保下一年度的预算和绩效奖金，以此化解唱片业对流媒体授权的阻抗。在历经瑞典首发、英国扩张及长达五年的准备后，Spotify 于 2011 年进入美国，现已发展为拥有超 7 亿月活跃用户和逾 3 亿付费订阅用户的平台。随着 Ek 在 1 月 1 日转任执行董事长，他将这套从瑞典研发验证再向英美拓展的地域扩张策略复制到了医疗领域。 面对医疗创新，Ek 明确表示相比被动投资，自己更倾向于亲自下场构建。他认为董事会席位往往面临尴尬悖论：创始人不听建议会令投资人焦虑，盲目听从则说明其缺乏领导力。因此，在 2018 年 Spotify 直接上市获得个人流动资金后，Ek 虽设立了 Prima Materia 支持前沿科研，但仍选择与联合创始人兼 CEO Hjalmar Nilsonne 共同创办 Neko Health，耗费五年时间进行静默软硬件自研，直至 2023 年才在瑞典开设诊所并逐步走向国际。 医疗体制错配：短期支付周期与预防脱节 (27:18) Ek 分析指出，美国医疗系统面临的根本挑战并非源于技术不足，而是根植于激励机制的系统性错位。目前美国医疗支出占其 GDP 的 18%，构成联邦预算中最大的单一支出项，其中仅心脏病一项的年度开支就超过了《财富》十强企业的年营收规模。尽管业内普遍认同慢性病是最大的医疗负担且在早期完全可控，但现行交付体系最初是为应对急性传染病而设计的，无法有效适配隐性、发展缓慢的慢性病。 援引查理·芒格关于激励机制决定结果的观点，Ek 强调美国私人保险与雇佣关系绑定的结构造成了短视的支付周期。美国员工平均每两到三年更换一次工作，导致医疗保险计划随之变更。对于商业保险机构而言，为用户支付 500 美元的早期筛查几乎没有直接的财务投资回报率（ROI），因为该筛查预防的可能是十五至二十年后的严重心血管疾病，届时患者早已转移至其他商业保险或 Medicare。这种结构性阻碍直接抑制了预防性筛查的大规模普及。 垂直整合早筛：Neko 的多模态数据飞轮 (12:35) 为了在缺乏数据支撑的传统体检中突围，Neko Health 推出了一项定价 499 美元、耗时一小时的全方位检测服务。客户在单次就诊中完成包含 53 项指标的现场验血，并步入配备高分辨率相机的全身光学扫描设备，采集超过 6,000 张微观图像对全身体表进行计算建档。针对人体平均拥有的 950 颗痣，系统通过连续的纵向追踪对比每年形态与纹理变化，协助驻场临床医生及皮肤科专家排查早期黑色素瘤风险，并同步采集动脉硬度、握力等心血管与生理基线数据。 Ek 强调，499 美元的定价并非赔本引流，而是依托软硬件全栈自研实现的正向单店经济效益。Neko 避开了传统医疗机构拼凑第三方硬件与授权软件的中间商溢价，从而使既有诊所实现盈利，并支撑其在纽约拉法叶街 300 号设立美国旗舰店，后续拓展至迈阿密和华盛顿特区。截至目前，Neko 历经八年研发已完成逾 10 万次扫描，连续三年发布同行评议数据，并推进了 10 项已完成或计划中的临床试验；实际数据表明，基础指标较差的人群在直观看到自身血管或皮肤健康图像后，反而展现出了最高的干预依从性。 技术治理：算力监管、开源与平台平衡 (40:00) 在探讨人工智能生态时，Ek 对科技界过度渲染末日论调提出批评，呼吁公众关注 AI 在降低诊疗成本、监测皮肤癌变等领域的现实价值。他认为技术历史始终在开放与封闭之间保持动态平衡，正如操作系统演进为 Windows 与 Linux、移动端演进为 iOS 与 Android 一样，开源模型与闭源前沿模型将在 AI 领域长期共存。在 Spotify 内部，工程团队既采用闭源大模型处理通用任务，也微调内部开源模型以降低推理成本并获取针对性性能。 针对 AI 监管框架，Ek 主张政策重心应从审查抽象的模型权重或静态浮点数门槛，转向监管物理层面的总算力集群。他认为在本地 PC 上运行开源模型的攻击能力极为有限，而调动 100,000 块 GPU 的实体则具备系统性破坏力，这类似于冷战时期对 Cray 超级计算机的硬件级管控。此外，谈及早期创业经历，Ek 回顾了在筹备 Spotify 前担任 Stardoll 临时 CTO 的经历，当时他通过重构架构将崩溃边缘的网页加载时间从四分钟压缩至一秒以内，促成了红杉资本的注资；在平台治理上，他主张在保留开放 RSS 标准分发的同时，通过专有平台功能维持产品迭代速度。 Daniel Ek 试图将消费级硬件的降本路径与多模态纵向数据模型引入医疗，以此打破传统医疗体系对慢性病防治的冷漠；但该模式能否在高度复杂的美国保险报销体制之外，仅凭自费早筛在更广泛的人群中实现长期的流行病学改善与宏观控费，仍有待未来数年跨国队列数据的进一步检验。

Detailed notes — zh

## Background 该视频由知名科技创投播客频道《All-In Podcast》于 2026 年 9 月 28 日在 YouTube 平台发布。节目简介显示，Spotify 联合创始人 Daniel Ek 做客对谈，深入探讨了创立 Spotify 早期击败盗版的幕后对赌、二次创业成立的预防性医疗公司 Neko Health 及其 499 美元全身扫描模式、美国医疗体系的激励扭曲，以及 AI 的开放与闭源之争、算力监管和早年执掌 Stardoll 的创业轶事。 ## TL;DR Daniel Ek 回顾了自己倾尽身家与唱片公司谈判、在盗版横行的瑞典艰难打磨出 Spotify 的早期历程，并阐述了如何将相同的商业扩张飞轮复刻到其创办的预防性医疗科技公司 Neko Health。他尖锐批判了美国医疗体制将巨额预算浪费在被动救治而非早期预防上的系统性激励错位，提出通过垂直整合软硬件将全套高精深度体检降至 499 美元，借助 AI 算法与纵向数据追踪防范重大慢性病与肿瘤。此外，他还深入探讨了开源大模型普及对于防止巨头垄断的重要性、应以算力规模而非算法代码作为监管锚点的政策建议，并分享了播客开放标准与早年重构虚拟形象社区 Stardoll 的幕后故事。 ## Atmosphere & style 整场对谈兼具老友叙旧的亲切松弛与顶级科技领袖深度复盘的思想厚度，氛围幽默而极具洞察力。四位‘Besties’主持人与 Daniel Ek 互动默契，频频就天使投资痛点、加州死板的 AI 立法以及全民‘自封各行专家’的舆论现象展开机智的调侃与辛辣吐槽。在欢声笑语的包裹下，Daniel 的表述始终沉稳、笃定且数据严密，从音乐版权泥潭到千疮百孔的医疗体制，真实生动地展现出一位硬核连续创业者面对复杂巨系统时迎难而上、用工程思维拆解死结的极客执着。 ## Bulletin - [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) Daniel Ek 复制 Spotify 从瑞典验证再进军英美的打法，创立深度医疗科技公司 Neko Health，耗时 5 年潜行研发后现已累计完成逾 10 万次深度扫描，并正式启动美国扩张。 - [15:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=900s) Neko Health 推出 499 美元的全套全身扫描与 53 项血液检测体检服务，利用定制硬件、6000 张高精照片与自研 AI 算法实现单体经济正向盈利，主打高性价比的年度健康投资。 - [27:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1620s) 美国医疗开支占 GDP 的 18%，但其体系脱胎于应对急性传染病的历史遗留，加之雇主医保捆绑导致的短期视角，使支付方严重缺乏动力为 10 至 20 年周期的预防性早筛买单。 - [40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) 在 AI 治理与技术路线方面，Daniel Ek 坚定支持开源与开放权重模型，主张监管核心应聚焦于物理算力集群规模而非纯软件代码，以打破科技巨头对算力与财富的垄断。 ## Key takeaways - [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) 创立 Spotify 初期，Daniel Ek 与合伙人用自身全部积蓄向瑞典唱片高管兜底次年年终奖与预算，成功打破了数字盗版泛滥下的版权授权僵局。 - [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) Neko Health 完整复刻了 Spotify 的市场扩张路径：先在瑞典本土深耕研发 5 年并验证产品，随后进军英国，最终进入美国市场。 - [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) Daniel Ek 认为顶级天使投资策略应是选对优秀的创始人后彻底放手，过度干预或盲从指令的董事会治理往往导致企业走向失败。 - [10:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=600s) 美国医疗支出占全国 GDP 的 18%，单中心血管疾病年开支即超数千亿美元；将医疗从被动救治转向低成本早筛是解决系统性危机的核心共识。 - [12:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=750s) Neko Health 借助智能手机产业链带来的廉价高精度传感器红利，构建多模态与纵向追踪数据飞轮，用类似 Spotify 推荐算法的逻辑预测未来健康风险。 - [15:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=900s) Neko Health 通过自研硬件全景扫描舱拍摄 6000 张超清照片，辅以 53 项血液即时检测与医生一对一复盘，在 1 小时内完成全套深度排查。 - [17:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1050s) 临床数据显示普通人平均长有多达 950 颗痣，传统肉眼面诊极易遗漏，而 AI 与计算机视觉的纵向比对能够精准捕捉年际间的微小病灶异常。 - [25:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1500s) 依靠全栈垂直整合自研硬件、传感器与算法，Neko Health 摆脱了昂贵的第三方采购与医保转诊补贴，仅凭 499 美元单次定价即可实现诊所自我盈利。 - [27:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1620s) 美国雇主与医保绑定的机制导致员工跳槽频繁，保险公司因无法获取 10 至 20 年后的远期财务回报，系统性缺乏支付预防性早筛费用的经济动力。 - [30:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1800s) 为验证预防性医疗的长期 ROI，Neko 采取将筛查成本门槛降至大众可承受区间的策略，并已推进 10 项临床试验以填补行业高质量结构化数据的空白。 - [33:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2010s) Neko 摒弃传统医疗器械十年一贯的研发模式，沿用科技消费品的敏捷理念，三年内迭代两代硬件基础设施，并打通与主流可穿戴设备的数据集成。 - [36:04](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2164s) 科技行业近年来过度任由末日论发酵，未能充分向公众宣扬 AI 在拯救生命与丰富人类生活体验方面的巨大正面价值。 - [38:40](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2320s) 模型推理成本已从每百万 Token 约 30 美元骤降至 13 美分，开源生态与低廉推理成本是防止价值向少数科技巨头过度集中的关键平权工具。 - [40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) 技术史证明开放与封闭将长期共存；针对 AI 的合规监管应当以大型数据中心的物理算力规模为核心准入维度，而非审查算法软件代码。 - [44:24](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2664s) Spotify 在播客业务上采取双轨策略：既利用站内专有功能提升互动变现，同时提供托管工具全面兼容开放 RSS 协议以实现全网分发。 - [47:11](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2831s) 在创办 Spotify 前，Daniel Ek 曾为 Stardoll 彻底重写后端技术架构，将页面加载时长从 4 分钟压缩至 1 秒以内，助力其成功拿下红杉资本注资。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) 创立 Spotify：用自掏腰包的对赌打破唱片业死局 节目开场，主持人回顾了 Daniel Ek 掌舵 Spotify 长达二十年的辉煌战绩：从零打造出拥有超过 7 亿月活跃用户、3 亿多付费订阅用户的流媒体帝国。就在今年 1 月 1 日，Daniel Ek 正式转任执行董事长。主持人 Jason 提起二十年前在某次互联网峰会上，Daniel 的联合创始人 Martin Lorentzon 探身向他展示 Spotify 的场景——在那个无人知晓订阅制为何物的年代，这一构想宛如天方夜谭。 谈及 Spotify 诞生的背景，Daniel Ek 强调那是 iPhone、YouTube 尚未问世，Facebook 仅限于大学校园的蛮荒时期。彼时全球音乐行业深陷数字盗版的泥潭，Napster、Kazaa 肆虐，美国唱片业协会（RIAA）甚至开始在全美四处起诉非法下载音乐的普通消费者。面对不可逆转的数字化浪潮，Daniel 和 Martin 坐下来头脑风暴，Martin 问他如果不考虑限制最想做什么，Daniel 坦言自己最热爱音乐，但去碰音乐行业“简直愚不可及”：复杂的版权法、严苛的唱片公司授权谈判足以劝退任何人。然而，在 Martin 连续不断的“如果……会怎样”的追问下，Daniel 终于构想出一个产品愿景——“让全世界的所有音乐触手可及”。 最初他们雄心勃勃地计划从第一天就推向全球，结果遭到了业内无情的嘲笑。但 Daniel 坦承自己运气不错，当时全球盗版最严重、最糟糕的音乐市场恰恰是他的祖国瑞典： - 瑞典早在 2000 年代初就普及了超高速宽带，用户可以随心所欲下载任何内容； - 苹果的 iTunes 根本没有进入瑞典市场，当地消费者毫无合法流媒体消费渠道可选； - 瑞典本土唱片业的总收入在盗版冲击下蒸发了近 80%。 面对绝境，Daniel 和 Martin 做出了一个孤注一掷的豪赌。他们掏出此前创业积累的全部积蓄，对唱片高管们开出了一个无法拒绝的条件：“我们保证你们拿到明年的年终奖，保证你们完成明年的财务预算。如果这事真像你们想的那么糟，一年后直接关停，但你们的奖金已经落袋为安；如果成了，你们也是赢家。”在长达数年的拉锯谈判后，唱片公司终于妥协。Spotify 于 2008 年底在瑞典上线即引发轰动，随后引爆英国市场，并在漫长的五年研发与博弈后，于 2011 年正式进军美国。 ### [05:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=300s) 复制 Spotify 剧本：从音乐版权到 Neko Health 的医疗革命 面对主持人关于“五年实验室研发才打入美国市场经历无数痛苦”的感慨，Daniel Ek 笑称自己正在重走当年的老路，甚至调侃自己或许是“痛阈极高”。他在 2018 年创立了专注于预防性医疗的深度科技公司 Neko Health，直到 2023 年才正式在瑞典推向市场，整整研发了五年；在瑞典验证模式后，Neko 复制当年的轨迹扩张至英国，如今也正式启动了进军美国市场的计划。这完全就是“医疗版”的 Spotify 扩张路径。 谈及从流媒体转向医疗的契机，Daniel 透露，早在 2012 或 2013 年接受《金融时报》（Financial Times）著名的“与 FT 共进午餐”采访时，他就公开大谈医疗健康问题。他认为世界上绝大多数事情都在沿着科技曲线向好发展，但医疗行业却是一个巨大的反例——投入越来越高，机制却严重失调。此外，Daniel 纠正了外界对他很早就涉足天使投资的误解：他在 Spotify 于 2018 年上市前根本没有多余的钱，完全是全身心扑在 Spotify 业务上长达十二三年。 上市之后，Daniel 虽涉足过少量投资，但他很快与现场的主持人产生强烈共鸣：自己骨子里深爱的是“从零打造（building）”而非单纯的资本运作。两位投资人对此深有感触地吐槽起董事会经历的痛苦： - 看着被投公司管理混乱却无法干预是极度折磨的； - 坐在董事会上给 CEO 提建议，如果对方不听，你会气得抓狂； - 如果 CEO 完全盲从并按你的指令行事，企业同样会完蛋，因为领航者失去了自主领导力。 最终他们得出了完全一致的共识：最好的天使投资策略莫过于找到极其出色的创始人/CEO，给他们开出支票，然后彻底放手不再干涉，往往能获得最好的结果。 ```mermaid flowchart TD accTitle: Daniel Ek 商业模式跨领域迁移路径 accDescr: 展示 Daniel Ek 如何将 Spotify 的市场验证路径完全复制到医疗公司 Neko Health 的发展历程中。 subgraph Spotify ["Spotify 历程 (2006-2011)"] S1["2006-2008: 瑞典本土技术研发与版权破局"] --> S2["2008: 瑞典本土爆火验证产品"] S2 --> S3["跨越海峡开拓英国市场"] S3 --> S4["2011: 历经5年攻坚正式登陆美国"] end subgraph Neko ["Neko Health 历程 (2018-至今)"] N1["2018-2023: 瑞典潜行研发5年医疗扫描方案"] --> N2["2023: 瑞典上线验证预防性医疗模式"] N2 --> N3["落地英国市场开启国际化扩张"] N3 --> N4["当前阶段: 启动全面进军美国市场"] end Spotify -.->|相同商业扩张剧本跨界迁移| Neko ``` ### [10:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=600s) 医疗体系的逆向激励与 Neko Health 的诞生动因 主持人与 Daniel Ek 探讨了究竟是什么促使他将目光从流媒体转向医疗健康领域。Daniel Ek 指出，想要真正改变一个领域，如果抱有极大的热情和行动力，最好的方式就是亲自下场去做。他首先给观众普及了一组触目惊心的宏观数据：美国目前将整整 18% 的 GDP 消耗在医疗保健上，这是联邦和全美预算中最大的一笔开支。更夸张的是，仅心血管疾病这一个单一疾病大类，每年在美国的开支就高达数千亿美元，这一数字甚至超过了全美《财富》前十强（Fortune 10）顶级公司的年营业收入总和。 面对如此庞大的死结，Daniel Ek 深入调研了整个医疗产业链，从制药研发效率一路研究到各类癌症。他最终发现，导致这一局面的根本症结在于慢性病与恶性肿瘤（如黑色素瘤等）的泛滥。令人扼腕的现实是，绝大多数慢性病如果在极早期被侦测出来，不仅完全可以预防，而且干预成本极低，患者本人、家庭及社区所承受的痛苦也微乎其微。“如果能早发现，一切都会迎刃而解。”整个医疗行业在此达成了唯一共识——必须将医疗体系从“患病后的被动救治（Reactive）”彻底转变为“前置的预防性医疗（Preventative）”。然而，真正引发巨大分歧的分水岭在于：具体该怎么做？ - **医疗开支占比**：美国医疗保健支出占全国 GDP 的 18%，是预算中最大单笔开销。 - **心脏病经济负担**：单一疾病群的年支出高达数千亿美元，规模超越全球前十大巨头的单体年收入。 - **早筛的核心价值**：早期介入不仅在临床上可控可逆，更能大幅削减后续昂贵的治疗成本与生命代价。 --- ### [12:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=750s) 借鉴 Spotify 数据飞轮：多模态、纵向数据与廉价传感器革命 Daniel Ek 与联合创始人 Hjalmar Nilsonne 坚信，变革的核心抓手在于“掌握更好、更全的数据”，从而形成改变医疗现状的正向数据飞轮。他们的终极目标是获取能够预测未来不良健康结果的前瞻性数据（Predictive Data），并在恶果发生前提前纠偏。为了实现这一预测能力，行业亟需两大维度的突破：其一是横跨多个模态（跨维度生理指标）的海量数据，其二是长期的纵向（Longitudinal）追踪数据。只有随着时间推移积累持续的时序数据，AI 预测的置信度才能质变提升。 关于实现的路径，Daniel Ek 敏锐地抓住了过去二十年智能手机产业链所催生的红利。智能手机的发展将大量微型元器件推向极致，催生了前所未有的廉价传感器与高集成度电子芯片。Neko Health 顺势而为，利用低成本高精度的工业级硬件，采集过去无法想象的体征细节。这一思路与 Daniel Ek 打造 Spotify 时的底层逻辑如出一辙：“这和我们在 Spotify 总结出的经验完全一样：我们收集到的数据越多，就越能精准预测系统下一首该为你播放什么歌曲。”从前期的机器学习到如今的大型语言模型与前沿 AI，海量数据喂养下的智能模型终于能够跨模态、纵向地对人体健康做出精准研判。 ```mermaid flowchart TD accTitle: Neko Health 预防性医疗数据飞轮 accDescr: 展示从硬件多模态采集到纵向数据积累，最终通过 AI 模型实现早期预测纠偏的闭环。 subgraph Hardware["硬件与数据采集层"] A["消费电子与传感器红利"] --> B["多模态健康指标采集"] end subgraph DataEngine["数据沉淀与分析层"] B --> C["纵向时间序列追踪\n(Longitudinal Data)"] C --> D["AI / 机器学习预测模型"] end subgraph Action["临床反馈与干预"] D --> E["提前发现潜在疾病"] E --> F["生活方式与临床低成本纠偏"] F -.->|"持续追踪校准"| C end ``` --- ### [15:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=900s) 重新发明体检体验：499 美元的“黄金一小时” 针对市场上此前关于早期观察数据能否真正改善健康终局的疑问，Daniel Ek 详细拆解了 Neko Health 目前推出的第二代（Gen 2）全套服务流程。该体检统一定价为 499 美元，是一套自底层垂直整合、完全端到端自主研发的颠覆性医疗体验，其使命被团队定义为：“打造你为自身健康所能投资的最具价值的一小时”。 用户走进诊所后，整个流程高效而严密： 1. **血液检测**：现场采集血液样本，即时化验多达 53 种关键血液生物标志物。 2. **全身皮肤扫描机（Skin Rig）**：受检者进入特制的全景扫描舱，由阵列相机系统瞬间拍摄超过 6,000 张超高分辨率身体照片，利用算法对全身每一处红斑、皮疹、病变以及哪怕最微小的痣建立数字索引档案。 3. **心血管与体能评估**：针对心脏功能、血液微循环系统进行深层次无创扫描，并搭配握力等经大量医学实证高度关联寿命与健康的经典体征测试。 4. **医师一对一长谈**：在整整一小时体检的尾声，受检者将拥有完全不受打扰的时间与执业临床医生面对面复盘，当场拿到包括血液检测在内的所有完整即时结果，并由医生进行详尽解读和定制改善建议。 该产品已在纽约全面落地铺开，并计划在接下来的数月内向全美各大核心城市快速扩张。 --- ### [17:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1050s) 临床实证：8 年研发沉淀、10 万次扫描与人体 950 颗痣的惊人发现 在被问及 Neko Health 与昔日硅谷备受瞩目的 One Medical 或 Forward 等主打会员制体检诊所的区别时，Daniel Ek 强调了 Neko 在底层技术与数据积累上的壁垒。Neko 并非跟风之作，而是早在 2018 年就已创立，迄今已在欧洲默默深耕研发了整整 8 年，累计完成了超过 100,000 次人体全身深度扫描。公司每年坚持向外界公开发布全量健康数据追踪调查报告，目前刚公布了第三年的随访成果。 数据揭示出令人振奋的临床事实：虽然相当一部分会员体征处于健康状态，主要聚焦于改善压力、饮食和睡眠等亚健康问题，但随访发现，那些初始健康状况最差的群体反而是获益最深、改善幅度最大的。Daniel Ek 解释道，当患者能极度直观地从高精图谱中亲眼目睹自己身体内部的各项数据与危机，并获得医生坐下来深入浅出地讲解病理隐患及干预方案时，这种视觉震撼与专业背书成了促使人们下定决心改变不良生活习惯的最大催化剂。通过极早期的排查，诊所成功捕捉到了大量此前从未被临床确诊的严重潜在疾患，帮助患者彻底逆转了健康走向。 谈及全身高清扫描的实际应用时，Daniel Ek 举了一个极具冲击力的皮肤科数据：“你知道普通人身上平均长了多少颗痣吗？在 Neko 的真实扫描数据统计中，平均每个人身上有多达 950 颗痣。”这一惊人数字引得现场主持人们发出不可思议的惊叹。若非依赖机器阵列与计算机视觉的全面建档追踪，任何传统门诊医生都绝不可能在肉眼常规检查中靠记忆监控数以千计的色素痣演变。 ### [20:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1200s) AI 与临床医生结合：Neko Health 的全身扫描与纵向追踪 在全身皮肤检测中，普通人身上常常分布着成百上千甚至数以万计的痣，仅凭肉眼根本无法逐一排查。Daniel Ek 指出，常规的医生面诊往往是被动式的：患者指着某颗感觉异样的痣，或者医生偶然注意到某处异常，再拿出检测设备并配合基础软件进行多频段成像，以此判断其大小与质感。而 Neko Health 彻底重构了这一检测流程：首先利用高精度 AI 系统扫描全身，自动标记所有潜在的风险病灶；随后由专业临床医生进行二次人工审核；若仍有疑点，则直接交由系统内常驻的多位资深皮肤科专科医生进行深度诊断。 ```mermaid flowchart LR accTitle: Neko Health 皮肤病变多级筛查流程 accDescr: 展现从全自动扫描到 AI 标记、临床医生初审、皮肤专家会诊及年度纵向数据对比的全流程。 subgraph S1["多级筛查管线"] A["高精度全身扫描"] --> B["AI 自动标记风险病灶"] B --> C["临床医生初审"] C --> D["常驻皮肤专科医生会诊"] end subgraph S2["纵向追踪"] A -.-> E["建立个人健康数字档案"] E --> F["跨年度微小异常变化对比"] end ``` 这种流程最核心的壁垒在于“纵向追踪（Longitudinal Tracking）”。在单次检查中，某颗痣可能未达到病变标准，但 Neko 会将其建立索引并建档。到第二年复查时，系统会自动比对两组数据，精准捕捉其是否发生微小但异常的形变或生长。Daniel 强调，即便是世界上最顶尖的医生，也绝不可能凭借人脑记忆记住患者身上成百上千颗痣一年间的细微变化。Neko 的愿景并非用机器取代人类，而是打造“AI 算法与顶尖临床医生无缝融合的医疗体验”。被问及具体职务时，Daniel 坦言自己并非 Neko 的全职 CEO，其联合创始人 Hjalmar Nilsonne 才是 Neko 真正的“大脑（brainchild）”与核心操盘手。 ### [22:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1350s) 499 美元单次体检：像定期看牙医一样投资健康 针对 Neko Health 定价 499 美元的全身扫描服务，播客主持人将其与市场上仅做血液检测的 Function Health（约 350-500 美元）及 Superpower 等做了对比。目前美国联邦财政预算中，医保（Healthcare）高居第二位，仅次于社会保障（Social Security），且债务利息支出也在迅速攀升。主持人算了一笔账：如果每位美国人每年或每隔几年做一次类似筛查，其成本仅占医疗总支出的 2% 到 8%，这是否能大幅削减未来的医疗赤字？ Daniel 表示，虽然公司对长期效益有自己的研判，但目前断言具体能为国家财政省下多少万亿支出还为时过早。不过从个人健康管理角度来看，Neko 强烈建议用户每年进行一次筛查。Daniel 打了一个生动的比方： > “我们大多数人每年都会雷打不动地去看牙医，但不知出于什么原因，绝大多数人却从不做年度深度体检。” Neko 的使命是为用户创造“投入产出比最高的一小时健康投资”。如果用户没有任何疑问且只关心核心体检数据，整个闭环耗时极短，极致高效。尽管 500 美元对许多家庭而言仍是一笔不小的开支，但面对潜在重大疾病的预防，这无疑是一笔极具性价比的健康投资。 ### [25:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1500s) 垂直整合的商业模型与 Daniel 的多重事业布局 面对主持人群体关于“500 美元是否会像传统医疗一样赔本赚吆喝、后续靠二次销售补贴”的商业模式拷问，Daniel 揭示了 Neko 的底层经济学原理。由于 Neko 采取了高度的“垂直整合（Vertically Integrated）”路径，从定制硬件、扫描舱、传感器到 AI 诊断软件皆为自研，极大地压缩了昂贵第三方设备的采购与授权成本。因此，在不依赖复杂医保转诊和高价附加服务的前提下，Neko 仅凭 499 美元的单次收费便已实现了正向的单体经济模型（Positive Unit Economics），目前旗下运营的线下诊所已经能够实现自我盈利。 在精力分配方面，Daniel 澄清自己并非只扑在 Neko 上： - **Spotify**：目前公司由两位联席 CEO（Co-CEOs）负责日常运营，Daniel 担任执行董事长（Executive Chairman），Spotify 依然是他倾注心血的“亲生骨肉”。 - **Neko Health**：聚焦全球最大的健康难题，投入大量精力提供战略支持。 - **Prima Materia**：与 Shakil Khan 共同创立的投资机构，旨在扮演欧洲科技创业者“所能找到的最佳联合创始人”，通过资本与经验反哺深科技生态。 ### [27:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1620s) 美国医疗困局：错位的支付激励与急性传染病遗留体制 对谈随后切入美国医疗制度的沉疴。主持人列举了美国人当前面临的三大生存成本压力：购房、教育与医疗。其中住房与教育成本尚可通过纠偏市场扭曲来解决，唯独医疗系统的荒诞令人发指——急诊缝几针收费 1.5 万美元；出厂成本 30 美元的药品卖到 2 万美元；年薪 20 万美元的医生仅仅面诊 8 分钟，保险账单却高达 6,000 美元。 被问及如果成为“一日独裁者”该如何手术式改造美国医疗时，Daniel 直言：“给我看结果，我就能告诉你背后的激励机制（Show me the outcome, I'll show you the incentive）。”他指出，现代医疗体系建立的核心逻辑脱胎于应对“急性传染病”的历史时期，所有的计费代码与赔付机制都是为“患者出现严重急性症状后进院救治”而设计的。 更致命的结构性断层在于“雇主医保绑定机制”： - 绝大多数美国人的健康保险与其全职工作深度捆绑。 - 现代职场中员工的平均跳槽周期只有两三年，一旦跳槽，背后的承保保险公司（Payer）也会随之更换。 - 这种短视机制导致保险公司根本没有动力为一个体征正常的年轻人支付 500 美元的预防性早筛；因为投入预防所带来的远期财务收益（如在 10 到 20 年后避免一场数十万美元的心血管搭桥或晚期肿瘤化疗）往往兑现在未来的另一家保险公司账上，当前承保方只会将早期筛查视作纯粹的当期成本支出。系统性的激励错位，最终让整套医疗制度在“等人生大病才肯花钱”的泥潭中越陷越深。 ### [30:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1800s) 医疗投资的回报周期与 Neko 的多模态纵向数据布局 谈及预防性医疗的商业与社会逻辑，Daniel Ek 深入剖析了这一领域最棘手的问题：如果一笔健康投资的回报周期长达 10 年、15 年甚至 20 年，该如何计算其投资回报率（ROI）？又该由谁来掏这笔钱？在现有的医疗体系中，无论是雇主还是商业保险机构，都不可能在缺乏可靠数据的前提下，盲目砸下数百万美元去押注 20 年后的预防效果。Neko Health 的破局策略首先在于通过工程化将筛查成本大幅压缩——将单次门槛降至可承受范围，即使回本周期长达一二十年，市场和机构的采纳意愿也会大幅提升。更深层次的支撑则在于构建“多模态纵向数据集”（multimodal longitudinal data sets），通过长周期的跟踪与对比，用硬核数据验证预防措施的真实临床有效性。为此，Neko 坚持每年公开发布健康统计报告，向公众和科研界展示其大样本数据的聚合成果。 ```mermaid flowchart TD accTitle: Neko Health 预防性医疗飞轮 accDescr: 展示 Neko Health 如何通过降低筛查成本、构建多模态纵向数据集、跨国对比及临床验证，推动早期预防医疗落地的闭环流程。 subgraph Cost["成本与采纳"] A["大幅降低全身体检成本"] --> B["提升个人与支付方采纳率"] end subgraph Data["数据与验证"] B --> C["沉淀多模态纵向数据<br/>(每年公开数据报告)"] C --> D["跨国人群对比<br/>(瑞典 vs 英国等)"] D --> E["推进高密度临床试验<br/>(已完成4项 / 进行中2项 / 筹备4项)"] end subgraph Impact["系统变革"] E --> F["将医疗系统可用数据量提升10-100倍"] F --> A end ``` 这种数据体系的另一大看点在于其天然的跨国属性。Jason 随之追问，在比较斯德哥尔摩与英国两地的人群健康数据时发现了什么差异，并根据常识打趣推测：瑞典人的身体素质应该比英国人好得多，而英国人可能有更多心血管疾病和糖尿病。Friedberg 与 Jason 随即探讨起这两大慢性病在医疗整体支出中的绝对主导地位。Ek 坦言，虽然 Neko 目前已完成了 10 万次扫描，但样本量对于直接在国家层面上对宏观健康结局下定论仍显为时过早。但他强调，Neko 正全力携手科研界推进临床验证，目前已完成 4 项临床试验，2 项正在进行中，另有 4 项即将启动。Ek 指出，科技从业者常常误以为医疗行业的数据极其庞大，但事实恰恰相反，现有体系内的高质量结构化数据少得可怜；如果 Neko 能够将可用数据量提升 10 倍乃至 100 倍，不仅是 Neko，整个医疗系统都将从中提炼出颠覆性的结论。 - **投资回报痛点**：传统模式要求机构在缺乏数据支撑的情况下，提前 10 到 20 年承担数百万美元的预防性支出；Neko 致力于将门槛降至数万美元甚至更低区间以激活采纳动力。 - **临床试验进展**：Neko 现已完成 4 项临床试验，正在进行 2 项，另有 4 项在规划中。 - **筛查规模**：已累计完成约 100,000 次全身深度扫描。 --- ### [33:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2010s) 可穿戴设备整合与 Neko 硬件基础设施的代际迭代 Jason 进一步将话题引向市面上火热的可穿戴设备（如 Oura、Whoop、Fitbit 以及 Apple Watch），并提到 Whoop 近期开始通过 Quest Diagnostics 提供血液化验服务。他指出，目前业内能测睡眠、心率和恢复指数，但从来没有人把长期血液指标与日常可穿戴数据全面打通。既然大部分手环类硬件底层 80% 的功能已经商品化（commodified），Neko 未来是会寻求生态合作，还是干脆自己推出一款每月 9.95 美元的专属硬件订阅？Daniel Ek 回应称，Neko 目前已经支持导入 Apple Health 数据，由于 Oura、Whoop 等主流品牌都能将体征写入 Apple Health，患者在 Neko 诊所面对医生时，临床医生已经能够结合日常连续监测数据与现场扫描指标，提供更为完整的 360 度全身健康评估。 与此同时，Ek 重点阐述了 Neko 与传统医疗器械公司截然不同的研发哲学。在传统的诊断行业中，标准做法是耗费数年进行实验室研发，随后在长达十年的时间里一成不变地销售同一款设备。而 Neko 沿袭了科技消费品的敏捷研发理念：从产品最初亮相到如今正式进军美国市场，短短三年内其硬件与扫描舱已经完成了两代基础设施迭代（Gen 2 infrastructure）。Ek 表示，当前推出的扫描服务仅仅是一个起点，团队的目标是不断将前沿的高价值诊断技术以亲民的价格打包塞入产品中，从而坚定践行其品牌核心使命——打造“为你个人健康花费的最有价值的一小时”（"the most valuable hour you can spend on your health"）。 --- ### [36:04](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2164s) AI 发展的节奏争议与科技行业对正面价值的宣传缺失 对话随后转向科技前沿最受瞩目的议题：整个行业当前是否应该人为控制 AI 的演进步伐（pacing the frontier）？对此，Daniel Ek 并没有直接给出非黑即白的立场，而是先用一段辛辣的社会观察展开调侃。他回顾道，过去几年公众与舆论的浮躁心态极其明显：“当疫情暴发时，我们所有人一夜之间都成了病毒学家；当第一艘油轮在霍尔木兹海峡或苏伊士运河搁浅时，我们瞬间全成了供应链专家；而乌克兰危机一爆发，大家又摇身一变成为了军事战略家。”这番话引发了全场大笑，Jason 顺势接梗：“你这是在当面点我们 All-In 播客吗？我们播客干脆改名叫‘All-In 实验室’算了，因为我们的观点太危险了，指不定哪天需要 Spotify 的播客机制来监管我们。” Ek 笑着澄清这并非针对某个人，但核心问题在于：许多人总是以极度自负的口吻断言技术发展，而事后往往证明连顶尖专家也是错的。在 Ek 看来，任何一项颠覆性技术都兼具极端的积极面与不可忽视的负面影响，关键在于当下人类如何主动将技术引导至理想的结局。他直言不讳地批评，科技行业近年来犯下的严重失误就是未能向大众充分展示 AI 的巨大正面价值，反而任由末日论与焦虑情绪被无限放大。以 Spotify 为例，AI 正在让音乐体验逼近终极形态——未来的产品不仅是生成一张歌单，而是能精准地为用户人生的每一个瞬间配上最契合的声轨，“让你在快乐时感到更纯粹的快乐，在悲伤沉溺时也能更深地宣泄情感”；而在医疗领域，Neko 依靠 AI 能够在单次检查中同时精准追踪身体上的 950 颗痣，防范致命的黑色素瘤。这些实实在在拯救生命、丰富体验的技术突破，才是 AI 最应被聚光灯照耀的价值所在。 --- ### [38:40](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2320s) 开源模型与成本骤降：打破 AI 算力与财富的垄断壁垒 在探讨 AI 未来生态的治理结构时，对话延伸到了开源 AI（open-source AI）与开放权重模型（open-weight models）的重要性。Daniel Ek 对开源生态表达了极为坚定的支持，并指出算力成本的断崖式下跌正在重构整个技术版图：行业内模型的输出推理成本已经从早期的每百万 Token 约 30 美元，暴跌至当前的 13 美分左右。这种成本的大幅优化具有不可估量的平权意义。 Ek 认为，开源模型与极低廉的推理成本相结合，能够彻底打破大模型技术被少数几家科技巨头垄断的风险，防止价值与财富过度集中于寡头之手。只有让前沿的模型能力以极低的成本辐射给各类广泛的企业、中小开发者以及普通个人，AI 真正的创造力与普惠红利才能够被全面释放出来。 - **推理成本剧降**：每百万 Token 的推理输出成本从 30 美元急剧压缩至 13 美分，降幅达两个数量级以上。 - **开源生态意义**：开放权重模型能有效防止技术壁垒和财富集中于少数科技巨头手中，加速全行业各细分场景的创新落地。 ### [40:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2400s) 开源与闭源的技术轮回：以算力而非模型本身作为监管核心 针对开源与开放权重（Open Weights）模型爆发所带来的监管风波，Daniel Ek 指出，技术历史始终在“开放”与“封闭”之间来回摆动。从早期的 Windows 对决 Linux，到移动互联网时代的 iOS 对决 Android，科技行业见证过无数次类似的交锋，而最终的结局往往不是非此即彼，而是两者并存。Daniel 强调自己向来支持开放模型，因为这能为整个生态系统注入巨大的创新动力。以 Spotify 自身为例，团队不仅大量采购并接入最前沿的商业闭源大模型（Frontier Models），同时也在内部微调了许多专属的开源模型。正是这种“两条腿走路”的策略赋予了 Spotify 更高的创新自由度，使他们能够绕开前沿模型高昂的推理成本，或解决通用大模型无法针对特定业务逻辑进行深度微调的痛点。 Daniel 提出了一个在当前政策辩论中常被忽视的切入点：为何大家不把焦点放在“算力（Compute）”的绝对规模上？在他看来，模型的威胁程度并不单由算法本身的智力水平决定，投入的物理算力规模才是更关键的衡量指标。如果某个机构动用 10 万张 GPU 来驱动模型，其破坏力与能力上限必然呈数量级放大；相反，一个普通人在家用电脑上本地运行开源模型，根本不可能发动毁灭性的网络攻击，特别是当防守方大企业拥有神话级（Mythos-class）前沿防御模型的时候。因此，将实际消耗的算力规模作为风险防范和准入评估的核心维度，远比去给纯软件代码发许可证更具可操作性。 主持人群对这一见解深表赞同，并借此讽刺了加州等立法机构在 AI 监管法案上的死板与拙劣。大伙打趣说，议员们当初试图单纯靠模型训练使用了多少 Teraflops 算力来立法，仿佛技术参数永远静止不变一样，甚至恨不得去机房数以太网线有几根、耗了多少瓦电来判定模型是不是“坏模型”。David Sacks 和 Chamath 补充指出，这与当年 Cray 超级计算机的管控如出一辙——当年 Cray 被视同核武器技术而受到严密管制，谁想采购都必须经过安全审查，因为人们害怕它被用来造核弹或破译密码；然而时至今日，当年超级计算机的算力甚至还不如厨房里的一台智能洗衣机。 - **开放与封闭的历史常态**：从 Windows vs. Linux、iOS vs. Android 到今日的大模型，技术生态最终皆走向双轨共存。 - **算力作为防御护城河**：在模型水平相当的前提下，掌控海量算力的一方天然具备压倒性的安全防御优势。 - **算力认证 vs. 软件审查**：相比审查不可控的开源算法代码，围绕大型数据中心的大规模 GPU 算力集群设立安全护栏与合规认证更为切合实际。 --- ### [44:24](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2664s) 播客生态的开放标准与 Spotify 的双轨平台策略 谈到播客（Podcasting）这一基于开放 RSS 协议发展起来的领域时，Jason Calacanis 抛出了一个尖锐问题：播客行业向来珍视开放标准，而 Spotify 近年来在播客领域大举扩张并打造了大量专属私有功能，外界始终关注 Spotify 将如何平衡开放生态与私有平台之间的关系。 Daniel Ek 解释道，Spotify 在播客领域的定位同样秉持了开放与封闭共存的逻辑。虽然 Spotify 平台自身开发了诸如站内专属评论区等深度交互功能，但许多人不知道的是，Spotify 旗下拥有强大的托管与聚合工具，每天都在帮助大量播客主播使用开放 RSS 标准，将内容一键同步分发到全网其他所有播客平台。Daniel 直言，如果创新必须完全受制于开放标准的制定，往往需要漫长的全行业协商与妥协，周期极其拖沓；因此 Spotify 的战略是“双轨并行”——既在自身平台上快速迭代专有体验，又全力兼容并维护开放分发标准。 ```mermaid flowchart TD accTitle: Spotify 播客生态双轨架构 accDescr: 展示 Spotify 如何同时支持开放 RSS 协议分发与站内专有闭环功能 subgraph OpenTrack["开放标准轨道 (Open Ecosystem)"] RSS["底层开放 RSS 协议"] Tools["Spotify 聚合与分发工具"] ExtPlatforms["外部第三方播客平台"] RSS --> Tools Tools --> ExtPlatforms end subgraph ProprietaryTrack["专有平台轨道 (Spotify Platform)"] App["Spotify 客户端"] Exclusive["专有交互功能 (互动评论/高级数据)"] Monetization["独家变现与受众运营"] App --> Exclusive App --> Monetization end Creator["播客创作者 (Podcaster)"] --> Tools Creator --> App ``` 听闻 Daniel 的表态，作为拥有 17 年播客资历的老兵，Jason 趁机以“地下独立播客老炮代表”的身份向 Daniel 提出功能需求（Feature Request），希望 Spotify 尽快支持新一代播客开放标准中的两项核心特性：一是直播通知功能（Live in RSS，即开播时通过 RSS 协议向 Spotify 页面实时推送信令）；二是打赏与捐赠通道（Value / Donations，允许听众通过 PayPal 等第三方渠道直接赞助主播）。Daniel 幽默地回应称，虽然现在 Spotify 规模庞大，很多产品决策已经不再由他一个人拍板，但他一定会把这两项诉求当面转达给核心产品团队。 --- ### [47:11](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2831s) Stardoll 早期创业传奇：从 4 分钟加载时间到红杉注资 话题转向了瑞典斯德哥尔摩最早获得美国顶级风险投资青睐的传奇互联网公司——Stardoll。Jason 提到 Daniel 在二十岁出头时曾在这家数字商品与虚拟形象订阅先驱公司担任技术负责人（CTO），希望他聊聊这段鲜为人知的早年经历。 Daniel 回忆，在创办 Spotify 之前，瑞典互联网泡沫时期的元老级创业者 Mattias 找到了他。这项业务最初起源于芬兰图尔库（Turku），是由一位老妇人和她的儿子创办的简陋网站，原名叫做“纸娃娃天堂”（Paper Doll Heaven）。后来欧洲知名风投机构 Index Ventures 发现了这个项目并收购了多数股权，打算将其彻底重塑并更名为 Stardoll。然而，创始家庭根本不懂如何将网站进行技术扩容（Scaling）。Index Ventures 的合伙人 Danny Rimer 极具说服力，极力游说当时已经铁了心要去创办 Spotify 的 Daniel，劝他先抽调 6 到 12 个月时间去芬兰救火。 当 Daniel 真正飞抵芬兰调阅系统时，被眼前的技术惨状彻底震惊：当时网站渲染一个简单页面的平均加载时间竟然高达 4 分钟，整个后端服务器已经完全被流量挤压到崩溃解体的边缘。Daniel 临危受命，对整个网站的底层架构进行了彻底的重构，并在斯德哥尔摩组建了一支全新的技术团队。经过全面重写，他们奇迹般地将网页加载时间从 4 分钟骤降到 1 秒以内。性能瓶颈打破后，全球流量呈爆炸式狂飙，随后成功吸引了硅谷传奇风投红杉资本（Sequoia Capital）的重金注资。Daniel 笑着感慨，完成这次技术交卷后，他向大家道谢告别：“好啦，我的任务完成了”，随后便全心投入创立了 Spotify。 - **前身起源**：芬兰图尔库由母子创办的网站 Paper Doll Heaven，后被 Index Ventures 控股重塑为 Stardoll。 - **技术危机**：被收购时服务器处于崩溃边缘，平均页面渲染时间长达 4 分钟。 - **重构成果**：Daniel Ek 组建新团队重写架构，将响应速度提至 1 秒以内，为公司换来了红杉资本的后续投资。 --- ### [50:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=3000s) Neko Health 的全美扩张版图与节目尾声 在对话收尾阶段，主持人们询问了 Daniel 全新医疗预防科技项目 Neko Health 的最新线下网点规划。Daniel 透露，Neko Health 在美国的首个全身体检中心正式设在纽约市 Lafayette 街 300 号（邻近东休斯顿街）。 尽管目前 Neko Health 仅在纽约有一处运营网点，且等待名单（Waitlist）排队极长，但其扩张步伐十分迅猛。团队正在紧锣密鼓地扩大纽约本地的服务中心数量，紧接着将在佛罗里达州迈阿密以及首都华盛顿特区（D.C.）开设全新体检中心。Daniel 表示，他们的目标是在未来 12 到 24 个月内将 Neko Health 的早期预防扫描网络铺向全美。Jason 赞叹其纽约选址极具情怀，正对着老牌摇滚胜地和历史地标，最后呼吁观众前往 nekohealth.com 申请候补名额，并在欢笑与致谢中结束了整场对谈。 ## Notable quotes > [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) 让全世界的所有音乐触手可及 — Daniel Ek > [00:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=0s) 我们保证你们拿到明年的年终奖，保证你们完成明年的财务预算。如果这事真像你们想的那么糟，一年后直接关停，但你们的奖金已经落袋为安；如果成了，你们也是赢家。 — Daniel Ek > [10:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=600s) 如果能早发现，一切都会迎刃而解。 — Daniel Ek > [12:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=750s) 这和我们在 Spotify 总结出的经验完全一样：我们收集到的数据越多，就越能精准预测系统下一首该为你播放什么歌曲。 — Daniel Ek > [15:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=900s) 打造你为自身健康所能投资的最具价值的一小时 — Daniel Ek > [17:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1050s) 你知道普通人身上平均长了多少颗痣吗？在 Neko 的真实扫描数据统计中，平均每个人身上有多达 950 颗痣。 — Daniel Ek > [22:30](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1350s) 我们大多数人每年都会雷打不动地去看牙医，但不知出于什么原因，绝大多数人却从不做年度深度体检。 — Daniel Ek > [27:00](https://www.youtube.com/watch?v=JEUboZzZGM4&t=1620s) 给我看结果，我就能告诉你背后的激励机制（Show me the outcome, I'll show you the incentive） — Daniel Ek > [36:04](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2164s) 当疫情暴发时，我们所有人一夜之间都成了病毒学家；当第一艘油轮在霍尔木兹海峡或苏伊士运河搁浅时，我们瞬间全成了供应链专家；而乌克兰危机一爆发，大家又摇身一变成为了军事战略家。 — Daniel Ek > [36:04](https://www.youtube.com/watch?v=JEUboZzZGM4&t=2164s) 让你在快乐时感到更纯粹的快乐，在悲伤沉溺时也能更深地宣泄情感 — Daniel Ek ## Mentioned Daniel Ek, Martin Lorentzon, Hjalmar Nilsonne, Jason Calacanis, Chamath Palihapitiya, David Sacks, David Friedberg, Shakil Khan, Danny Rimer, Spotify, Neko Health, Prima Materia, Stardoll, Index Ventures, Sequoia Capital, Napster, Kazaa, RIAA, Apple, iTunes, Apple Health, Apple Watch, Oura, Whoop, Fitbit, Quest Diagnostics, Function Health, Superpower, One Medical, Forward, Financial Times, Cray Supercomputer ## Video description (0:00) Daniel Ek joins the Besties! (1:14) Starting Spotify: Beating piracy and betting his own money on the record labels (6:10) Neko Health: The Spotify playbook for healthcare, the $499 body scan, and why US healthcare waits until you're sick (36:04) AI: Tech's failure to sell the upside, open vs. closed models, and regulating compute (44:24) Back to Spotify: Podcasting's open standards and the Stardoll origin story Thanks to our partners for making this possible! As wealth grows, complexity compounds. Between investments, tax strategy and estate planning, coordinating it all can feel like a full-time job. Creative Planning can help with the heavy lifting. https://ad.doubleclick.net/ddm/trackclk/N2598215.3565131ALLIN/B36182939.452362736;dc_trk_aid=646454825;dc_trk_cid=260464105;dc_lat=;dc_rdid=;tag_for_child_directed_treatment=;tfua=;ltd=;dc_tdv=1 Starting a business? Northwest Registered Agent gives you everything you need to build a complete Business Identity including free tools and built-in privacy. Get more at https://www.northwestregisteredagent.com/ALLINFREE Follow the besties: https://x.com/chamath https://x.com/Jason https://x.com/DavidSacks https://x.com/friedberg Follow on X: https://x.com/theallinpod Follow on Instagram: https://www.instagram.com/theallinpod Follow on TikTok: https://www.tiktok.com/@allin Follow on LinkedIn: https://www.linkedin.com/company/allinpod Intro Music Credit: https://rb.gy/tppkzl https://x.com/yung_spielburg Intro Video Credit: https://x.com/TheZachEffect #allin #tech #news [Source](https://www.youtube.com/watch?v=JEUboZzZGM4)

Source quotes

[0:00] Daniel Ek: 让全世界的所有音乐触手可及 [0:00] Daniel Ek: 我们保证你们拿到明年的年终奖，保证你们完成明年的财务预算。如果这事真像你们想的那么糟，一年后直接关停，但你们的奖金已经落袋为安；如果成了，你们也是赢家。 [10:00] Daniel Ek: 如果能早发现，一切都会迎刃而解。 [12:30] Daniel Ek: 这和我们在 Spotify 总结出的经验完全一样：我们收集到的数据越多，就越能精准预测系统下一首该为你播放什么歌曲。 [15:00] Daniel Ek: 打造你为自身健康所能投资的最具价值的一小时 [17:30] Daniel Ek: 你知道普通人身上平均长了多少颗痣吗？在 Neko 的真实扫描数据统计中，平均每个人身上有多达 950 颗痣。 [22:30] Daniel Ek: 我们大多数人每年都会雷打不动地去看牙医，但不知出于什么原因，绝大多数人却从不做年度深度体检。 [27:00] Daniel Ek: 给我看结果，我就能告诉你背后的激励机制（Show me the outcome, I'll show you the incentive） [36:04] Daniel Ek: 当疫情暴发时，我们所有人一夜之间都成了病毒学家；当第一艘油轮在霍尔木兹海峡或苏伊士运河搁浅时，我们瞬间全成了供应链专家；而乌克兰危机一爆发，大家又摇身一变成为了军事战略家。 [36:04] Daniel Ek: 让你在快乐时感到更纯粹的快乐，在悲伤沉溺时也能更深地宣泄情感

Source quotes

[1:14] Daniel Ek: there's no way to put the genie back in the bottle. [1:14] Daniel Ek: a really stupid idea [1:14] Daniel Ek: it would feel like you had all the world's music at your fingertips. [5:00] Daniel Ek: If this really is as bad as you think it is, then you just shut it down after a year, but you still made the bonus. And if this actually works, you win as well. So you win no matter what. [12:35] Daniel Ek: the most valuable hour you can invest in your health. [27:18] Charlie Munger: show me the outcome, I'll show you the incentive [35:00] Daniel Ek: We're walking into COVID and every one of us became a virologist. We were all of a sudden supply chain experts after the first oil tanker in the Hormuz Strait got stuck somewhere. And then with the Ukraine war breaking out, we were all military experts and military strategists. [35:00] Daniel Ek: terrible disservice [35:00] Daniel Ek: feel more

Original transcript

[0:00] All right, everybody, welcome. back to the all in interview show where we the all in podcast dedicate an hour to Just some of the great thinkers, creators of our time. And today will be no different. Daniel Eck is with us. You know him, the co-founder of Spotify. which he started 20 years ago, and he ran it for those two decades to an extraordinary state of affairs today. well over 700 million active users. over 300 million premium subscribers, but On January 1st this year, he shifted and became the executive chairman with me. and David Freeberg today. Daniel Mack, how are you sir? Welcome to the program. Well, thank you so much for having me. It's good to be on. Yeah, it's good to see you. We've known each other for 20 years. I remember the first Shack, your partner, leaned into me at an Internet summit or something and said, hey, check this out. And he showed me Spotify. But this was at a time when... uh nobody had a subscription service so we're going to talk today about your new startup which is incredible in the healthcare space but take us back to the launch of spotify and and how that crazy idea got started and then how hard it was to break through with the music industry which are also known as the hardest partners in the world. Go to creativeplanning.com/allin. This is pre-iPhone, pre-I mean, Facebook existed, but it was college only. YouTube didn't exist. So just like that kind of dates me. But this is before all the things we now take for granted. And, um... And the world at that time, and the music industry at that time, was in free fall. Because music piracy was kind of rampant. So we had Napster, we had Kazaa, we had all these things. And there was an organization called the RIAA that went around and actually sued individual consumers in the US for illegally downloading music. And so I was sitting in Sweden and I was thinking to myself, you know, this is wild and this is crazy. There's no way to put the genie back in the bottle. Me and my co-founder, Martin, we were sitting and spitballing startup ideas at the time. Martin asked me, "If you could do anything you'd like to do, what would you do?" I said, "My real passion is music, but that seems like a really stupid idea." He said, "Why is that a stupid idea?" I told him, "It's really hard because you have to do all these deals with the record companies. You have to then figure out how to do licensing and copywriting." ride and you have to do all of these different things. And I said, well... What if you did this? And then he kept asking these what if questions. And eventually I ran out of reasons to say it. So I said, well, I guess you could do something. And if you did something, it would feel like you had all the world's music at your fingertips. And he's like, well, that sounds pretty good. And I said, well, okay, well, let's give it a try. We started trying to license music. Originally, the idea was to try to go global from day one, and everyone was just laughing at us when we were originally saying that. But I sort of lucked out because it turned out that one of the worst music markets in the world was my home country, Sweden. And part of the reason why it was one of the world's music markets was because we had super fast broadband already from the early 2000s. that they could download pretty much whatever they wanted. So they were doing it. And iTunes actually wasn't even available in the marketplace. So there was literally no legal options. So I think the music industry had lost like 80% or something of its revenues. And so it was like a... It was kind of this crazy of, you know, coincidence and kind of all of these things, the conflicts [5:00] where We figured out, you know, that me and my partner, we basically took all of our own money that we had made from prior startups. And we said to the record labels, what if we would guarantee you that basically you'd keep your bonus for the next year and you'll make your budget for next year? You know, if this really is as bad as you think it is, then you just shut it down after a year, but you still made the bonus. And if this actually works, you win as well. So you win no matter what. okay, well, this seems kind of like an interesting idea. And eventually, after a few years of negotiation, agreed to it. So we launched in late 2008. And it instantly became a huge hit. We then took it to the UK and made it a huge hit there. And then in 2011... 15 years ago now, we launched in the US. And yeah, I mean, and the rest is sort of a history. I mean, isn't that amazing, Freiburg? Five years in the laboratory just to get to the US market is quite a journey. That's a lot of pain and suffering. Yeah, but you know, the crazy thing is I'm on the same journey now again. Yeah, here we are. Neko actually, as crazy as it is, we started the company eight years ago in 2018. And we started it in Sweden in 2023. So it took us five years to develop the product, try it out in Sweden. And then once we've proven it in Sweden, we took it to the UK and did it there. And then from there on, now we're launching in the US. So it's an exact Spotify playbook again, but this time in healthcare. One of the afflictions of great entrepreneurs is you're entrepreneurial, right? I mean, you just can't stop. Or maybe I just have a high pain tolerance. I'm not sure which one. Yeah. Once things got smooth, you had to get back into it. Yeah. At what point did you start thinking about doing other things, Daniel? Were you actively angel investing at Spotify over the years as CEO? The business is building hundreds of millions of users, public, successful. Things are growing nicely. You've got a great leadership team. We've obviously met a lot of your team over the years. How did you start to do these other projects? I know you've got a few others and focus on Neko. Neko. Yeah, it's kind of crazy. I'd forgotten about it. But but my team now showed me very recently, like I apparently did this lunch with the FT interview in, I think 2012 or 2013 or something. just after the US launch of Spotify, where I was talking about healthcare already back then. So I've been thinking about healthcare specifically for a very, very long time. Um, and it's just been, you know, I, I think a lot of people just talk about sort of all the stuff that's going bad in the world. But my view is that actually most things are going pretty well and, uh, it's kind of up and to the right on the technology curve and all the things that we're getting. But I try to focus on the problems that aren't going that well. Um, you know, one of those happens to be healthcare. So I've been kind of interested about this as like, you know, how is it that we're spending and what can we do about it? And so that really kind of got me into the space. I think health care is kind of unique and kind of special. I honestly didn't have any money until Spotify went public, so I wasn't a prolific angel investor. What year was that? That was 2018 we went public. 2018, right. Yeah. So I... I probably started doing a little bit of investing in 2018, but I had no money before then. So, you know, it was literally just kind of all in focus on Spotify for, I guess, you know, the better part of 12, 13 years before I made any investing in anything else. And then, you know, you and I, we were talking about it at this dinner like a week ago or something. But, you know, I then tried to do a little bit of investing. I'd actually already started Netco at that time. And I kind of just realized that I love building way more than I loved investing. And so right now I came to the same conclusion. It's so painful to watch companies you invest in. This was our whole dinner conversation. Get mismanaged and you can't do it. time on boards. That's why I don't join boards anymore. You're sitting on a board, you advise a CEO and they don't listen to you and you pull your hair out like, I would have done something different. Why didn't you do this? And And if they are listening to you and you're effectively instructing them on what to do, it also doesn't work because they're not leading. And so I became... Pretty committed to this idea. You find amazing entrepreneur, founder, CEOs. Give them money and never look at it again, and you'll be fine. Generally better than not. [10:00] And then, you know, the flip side is, If that's really what you're passionate about and you have the activity or the agency, it's to step in and do it yourself. Yeah, the more instruction they need, probably the less chance they're going to succeed. So, Daniel, maybe you could just give us the overview on NetCo, kind of what the business – Yeah, and how you came to that idea, right? Yeah, how did you come to this specific idea? I started researching the space. a very, very long time ago. And maybe just as a quick primer for the audience, the U.S. is spending 18% of its GDP on healthcare. It's the single largest line item in the budget right now. And if you take heart disease alone... It's in the hundreds of billions of dollars that are being spent on that disease group alone in the US alone. I mean, it's just insane. It's larger than fortune 10 companies revenues, just one disease group. So it's a gigantic problem, as I mentioned. And so I started looking into everything from like, you know, drugs to, you know, can you make more efficient drugs? Is it about cancers? And and basically started like targeting, like, what are the real reasons why it looks that way? And I think that the the real conclusion ended up being there's lots of reasons, not one. You know, we see but we see a few really big ones. cancer, melanomas being another big one. So there's a few sort of major groups of illnesses that is over representative. But the major one is really chronic disease overall. And The crazy thing about chronic disease when you read about it is that if you discover these things early, they're totally preventable. And the cost of dealing with them is very small and the suffering for the person, you know, that are having it and their family and community is very small, too. So the key thing is if you discover things early, you're actually good. And so then the question is, well, why don't we discover things early? And it happens to be that we don't have enough data. So everyone in the entire health care industry is in agreement that we have to take health care from reactive to preventative health. And but then, you know, there now all the disagreement starts, which is how do you do that? And and our view, my co-founder Yelmer and I, our view is is really that it all starts with having better data. this virtuous flywheel of them changing healthcare. So predictive data that allows you to be predictive about future adverse health outcomes and course correct ahead of time. Yeah, that's that's the holy grail is to get to that point. And and essentially our view is that the way you get there is that you need a lot more data than you have today, both from each modality that you gather data, but also across multiple modalities. And you need it longitudinally. So you need it over time. more data and you had it over time, you'd be able to have a lot higher likelihood of doing those kind of predictive analysis. And then what we then said is, well, okay, how do we do that? And it so happened to be that if you think about really what's happened in the last 20 years, one of the biggest innovations, of course, is the smartphones. And we've had more and more sensors packed into those smartphones. So there's more and more cheap sensors and cheap electronics that's available to us. So our view was like, okay, well, we should be able to utilize this to build a lot more sensors than what's been able to be built before. free the LLMs and everything like that, but it was then called machine learning. But today we talk about it as AI. We should have a lot better intelligence. So if we have all these massive data sets, we should be able to make really good predictions about it. And obviously, again, similar to that, this is exactly what we learned with Spotify, too. The more data we're able to gather, the better predictions we're able to give you about what song to play next. in healthcare, but obviously it's, across multiple modalities and now across longitudinal data sets as well. Is there evidence today that some of this early observational data can affect ultimate health outcomes? I mean, I can see this working for metabolic conditions, things related like exercise depletion rates and... blood sugar and sleep, but ultimately the correction there is [15:00] typically diet, exercise, What are some other examples of major diseases that we can kind of identify ahead of the curve Course correct. And maybe just explaining what Nico is for people who don't know. Yeah, yeah. Maybe I'll start there and then we can get into it. So in a very simple way, the experience of NECO today and in Gen 2 is it's $499. It is a reinvented healthcare experience from the ground up, vertically integrated. It's really kind of an end-to-end solution that we've designed from the GradAlp. And we designed it with the goal of creating the most valuable hour you can invest in your health. That's the mission that we have. And that's what we're coming to work to do every day really at work. The experience is really you come in. The first thing you typically tend to do is that you draw blood from us. 53 blood markers. You then go through our skin rig where we have a camera system that takes over 6,000 super high-res images across your body where we index every mold and lesion and all of the possible rashes and rednesses you may have as well. That gets indexed. We then look into your heart and blood circulation. We look at grip strength, all those sort of traditional markers that we have very good scientific evidence are a valuable indicator as you think about sort of health. And then at the end of this visit that lasts for about an hour, you get uninterrupted time with a clinician and you go through all of the results, including the blood works during that hour. well when it comes to your health. So that's the experience that we're launching with today in New York, in the US, and obviously we'll launch in many more places across the United States over the next coming months. And just the identification of things ahead of the curve, because I've heard of other there was one medical, there was forward. I mean, there were a few other kind of attempts at making these integrated centers where you go in and get preventative checkups with their own instrumentation. And I think I've heard the thesis a little bit before. What have you identified that may be novel, Daniel, that we haven't seen in the past? I think first and foremost, the important thing to mention is, as I mentioned, we started this in 2018. So this company has been around for eight years. We've done more than 100,000 scans already. So we actually publish data every year. So we've just published our third year data survey where we go through the health outcomes that we have across our members. underlying medical situation that is undiagnosed. that we discovered. So, you know, again, the good news is many of our members are healthy and we're just talking about improving their current health status. And as you correctly pointed out, a lot of that ends up coming around stress, around diet, around sleep and those things. But the most encouraging thing we're finding now as we've done these studies for three years in a row is that the people with the worst health status, are the ones that are improving the most. So it turns out that this ability of having a clear visual understanding of where you are in your health journey currently today, and having a clinician that sits down with you explaining the situation and possible remedies to that is a huge benefactor for getting people to start improving things around lifestyle. So we've seen of them having to do this NECO and prioritizing their health. We've obviously seen people with severe medical underlying illnesses that have gone from discovering those and then treating them and having much better... overall health as a result. So it's really kind of all across the basis, but I'll give you a very concrete example. So take dermatology as a great example. The average... Person at Neko today has 950 moles. It's pretty crazy. The average person has 950 moles. Yes. That's insane. [20:00] You can't tell, I guess, is the punchline here. They're like not detectable by the human eye. Well, if you really went through your skin, you'd be surprised how many you'd find. I mean, it's a big dispersion. Some have, you know, many, many thousands and some of them have 500. Seems like a perfect AI application for AI, obviously. Exactly. So if you think about it across a normal doctor's visit, how likely is it that So normally what happens in a normal doctor visit would be you may say, look, I have a mole or the doctor sees, hey, let's take a look at that mole. It looks a little bit funky. And then they bring out their their tool where they inspect it. They usually these days have a software that checks against whether this mole, they take a camera across many different areas. frequencies to then discover both the the size of this mole but also the texture of the mole and so on and so forth we literally do that and we first have an ai system that flag possible risk factors and then we have a human clinician that then goes through and reviews those as well and should we even at that point find something we then have expert dermatologists on staff multiple of them of like what happens in one visit. But let's now presume the scenario that we'd have multiple visits. And maybe one year, it's actually, we don't discover anything. But because we've cataloged and indexed every single one, we can actually now, in a longitudinal base, look at them from one year to the next year. So it may be growing abnormally. These are things that if you think about it, even the best doctors in the world can't possibly remember how your mole may to another. So again, back to your point, this is a perfect application of AI, right? But we look at this as like, this is a perfect application of how healthcare should look like. It should seamlessly be both AI and amazing clinicians. in a great package experience. So that's essentially what we're trying to do with NetGo. That's great. And are you actively running it as CEO, Daniel, or are you operating part time? Yeah, no. So similar to Spotify these days where I have two great CEOs, I have an amazing co-founder and partner, Yalmer, who's really the brainchild of Neko. And he's the one who's doing most of the work, even though I'm sitting here and taking most of the credits at this point. But yeah, he's the real brain. You know what I'm struck by? The cost of this is so cheap, $499. You also have Function Health and Superpower, two other great services, about $350, $400. They just do blood work. But if we look at the spending, right now, it's healthcare number two, social security number one, and fast, your favorite coming up is the interest payments. 4% or 8% of the cost of the total we spend on healthcare. I'm curious, is this something you expect people to do yearly or something every other year? What impact do you think an investment of every American getting this done every year or every other year, or maybe every three years even, which would be down to 2% of our budget, what impact would that have on the forward-looking budget and the forward-looking spend? Have you done those kind of calculations? Yeah, I mean, look, um, We obviously have a view, but I think it's too early to say. So I don't want to sit here and say that we know conclusively what will happen. But I think that what we do recommend our members is to try to do this on an annual basis. And the way I kind of make this analogy is most of us today are trying to go to a dentist on a yearly basis. For some reason, we don't do an annual health checkup, most of us, on an annual basis. So for me, this should be as... as regular as going to the dentist or doing anything else like and it should be an amazing experience this should be um you know again as i said what we're really doing as a mission here is is trying to create the most valuable hour you can invest in your health and and you know we know a lot of people are busy so we want to be brutally efficient with people's time so you can actually if you don't have any questions and you're just interested in really doing the test results you it. So we're not talking about a lot of time to do that. And obviously, $500 is still a lot of money for people. But I do think it's a great investment if you can afford it to do. Starting a business usually means juggling a bunch of different services [25:00] just to look legit. Not anymore. With Northwest Registered Agent, you get a complete business identity all in one place. A business address, domain, website, phone number, and more. With privacy protection built in from day one. Don't spend hundreds or thousands on services you can get from Northwest for free. Visit northwestregisteredagent.com slash all in free and $500 you lose money on every visit or break even and then like there's some profit that comes from maybe the upsells and take us to the economic model here and then how that will work. with the insurance industrial complex as dysfunctional as it is here in America. Yeah, well, so the benefit of being vertically integrated is that we've been able to build, you know, everything ourselves, which means we can also cut costs in a pretty dramatic way. So within that $500 a year, we are the unit economics are positive. And we have clinics today that are profitable that we're operating already today. It's a great value for people. But because we do all the things ourselves, we're able to still make this economically viable for us to create a great business and to be able to grow and invest in that business so that we can spread it to more places as well. So Daniel, are you focused exclusively on this project? You have other stuff you're working on? I'm not exclusively focused on it, but it's obviously one of the things that are taking a lot of my time, and rightfully so. It's one of the biggest problems in the world. But I'm still spending time with Spotify and Exec Share there. So that's still, you know, my baby, even though I'm not involved in it on an everyday basis. I created together with Shaq that you mentioned earlier, where we're trying to really do um you know what i mentioned before which is like how do we how do we pay it forward and kind of almost be like the greatest co-founder you could possibly find that's that's what the ambition of primo material The healthcare thing really perplexes me. I kind of look at the biggest issues for affordability for people in the U.S. right now. It's buying a home. dealing with your educational costs, and then healthcare costs is the top three. I feel like we can solve education costs and we can solve home costs with certain market incentives that have been distorted. But healthcare is much more complex. As you look at the U.S. system, Beyond just kind of preventative care, the cost for servicing patients in a hospital, like why is it $15,000 to get stitches at an ER? You know, why is it $20,000 to get a drug that costs 30 bucks to make? You know, why when the doctor is making 200 grand a year, does the doctor seeing me for eight minutes? cost my insurance company $6,000. I mean, what is going on, from your view, that extends beyond like get in front of all of the catch up problems and so on. But what's structurally going on with the healthcare system in the US? And if you were emperor of the U.S. for a day and you got to say you could do anything you wanted, what are the top kind of three things that were missing that maybe... Neko doesn't necessarily address today, but are kind of things that you would recommend we fix? Well, I mean, again, I'm definitely not an expert on the U.S. health care system. But for me, it is like the show me the outcome. I'll show you the incentive. um it's it's one of those classical problems and if you really really think about it at the moment the entire system uh is predicated on and the health care system was built around the time when we were dealing with the infectious disease that was how the health care system was built uh and and dealt with so all the incentives are really around that which means we're fixing you acutely when there's massive amounts of symptoms so where we have to go is we have to go to the That means it has to be way more long term today. So I'll just mention one of the problems that exists is because normally your health care is tied to your employment in one shape or form. And because the average tenure of your employment is not that very long, it means that you may have only two or three years at one employer and then you switch. And when you switch, you switch insurers too. or twenty years to do that. Those are just some of those sort of incentive things like how do we [30:00] How do we look at the ROI of some investments that may be 10, 15, 20 years into the future? And who takes those investments? These are not easy... things to answer and I don't claim to have all the answers to them. But one of the things we're trying to do in that is we're just trying to bring the cost down so that the ROI doesn't have to be that we have to invest or the insurance industry have to invest or employers have to invest in millions of dollars speculatively hoping to get it back in 20 years time with very little data. So our view is like if we lower it, so it may not be a million dollars, it may be, you know, I make it up tens of thousands of dollars, even if that payback time was 10 or 20 years, I think that there will be a higher propensity and likelihood for a pickup on that. And then the second thing we're trying to do, obviously, is we're trying to add more data to the system. likely with these multimodal longitudinal data sets to be able to kind of actually see, well, what's the efficacy of these things over time. This is also part of the reason, by the way, why we are investing in releasing our data every year of what we're finding so that people can see, okay, well, what happens both on an aggregate NECO population, which is also super cool because Right. Like and so this is one of the few health care companies that are it's not global, but at least multi country systems straight away. So that's also super interesting. Like, what are we finding in the UK that may or may be similar or dissimilar to the US? That's going to be super. What did you find between Stockholm and the UK? I mean, I can take some guesses, but I'm guessing people in Australia. Sweden were much more fit in the UK. You probably had more cardiovascular disease and diabetes. Yeah. And those are the two, I believe, you tell me if I'm right, Friedberg. Aren't those like the two biggest spends in terms of health care spending? I think diabetes is number one. I could be wrong. But cardiovascular diabetes seems to be the big two. So what did you learn between the two populations? Well, I mean, it's really early and it's still on relatively small base. It's 100,000 scans, as I mentioned before. But I don't know that I think we're ready to speak about sort of population level health outcomes. But as we grow in NECO, I think we will. And that's going to be super cool. And then, you know, we're already today doing lots of clinical trials. We've done four already. We have two underway. We have another four we're going. research community as well deeply around these data sets that we already have today and that we will gather more on tomorrow too and I think my point with all of this is really just to say look we actually it's for people in tech the amount of data that exists in the healthcare system is actually not as the data sets are not that large that's the crazy part and so you know I think it's or 100x the amount of data that becomes available, we're going to be able to draw super interesting conclusions about that, not just Netco, but the whole healthcare system. That's really the bet we're making. Putting the wearables in there is going to be really interesting with the Auras, the Whoops, the Fitbits. You start with Apple Watches. You start correlating that data with the blood work. I noticed... whoop offering labs now through quest and i don't think maybe you can uh sync you know your uh borrow whoop with some of those but we don't i've never seen any data you know we have sleep we have recovery we have heart rate you know but i don't think anybody's put the blood work together with that And then your system, I mean, this could really... change everything in terms of healthcare if we can correlate that. Maybe you could speak to the impact you think wearables are going to have and Do you think you partner with them or you just have a NECO wearable? You eventually just make part of this subscription at $9.95 because it does seem like there's These are commodified for the basic 80% of it, I'm sure. So what are your thoughts there? Well, already today we do allow you to import your Apple health data. So if you do have wearable data, We will take that. So anyone who and a lot of the big or whoop, et cetera, do allow you to write a lot of that to Apple Health as well. That's just one one sort of example of that. So, yeah, we would we would love to have, of course, more data because that enables us to, again, when the clinician sits down with you to be able to give you a much more 360 perspective about your current health status. that we encourage. But I do want to say also that I think that this is just like [35:00] The service that we're launching today, it is very rare in itself that when you create a diagnostic product, the normal way of doing that in this space is you spend X amount of years doing R&D for that thing. Then you literally spend the next 10 years selling exactly that thing. And with Netco, within this three year when we launched and now launching in the US, we're on Gen 2. infrastructure. And so I think what you should expect with Netco is that we're launching with one thing today, but we're going to keep adding more and more and more value and we're going to keep adding more and more, you know, valuable information. diagnostics that we can find and we can put in at the price point. So that's kind of the goal. So this is just the beginning. And the goal is obviously to make this even more valuable so that we can deliver against our mission of having the most valuable hour you can spend on your health. Daniel, do you want to take a step back and talk about what's going on broadly in the tech industry right now and AI? What's your view on pacing the front here? I want to hear these questions. Yes. Do we need to pace the frontier? I'm struck by the fact that if we kind of look at the world, and certainly over the last few years, I kind of always say, you know, we're walking into COVID and every one of us became a virologist. We were all of a sudden supply chain experts after the first, you know, oil tanker in the Hormuz Strait got stuck somewhere. And then with Ukraine war broken out, we were all military experts and military strategies. Are you talking about the All In podcast right now? I don't know. I'm just talking. It's actually – we renamed it the All In Laboratory Pod. It's the All In lab because we are just – We're just a lab. We need to be regulated because our opinions are so dangerous, you know, that maybe Spotify's podcasting could regulate us. Yeah, my point is just like everyone expresses these things with such extreme confidence. And it turns out most of the time we ended up being completely wrong, even the biggest experts. for how to use these technologies. And, you know, my view is that every great technology has extreme positives and extreme negatives and is really up to us now to steer this towards the outcome that we want. I think it's great that the debates are being had, but honestly, I have no sort of strong opinion one way or the other if we should pace it or if we should not. I just know that if I look at health care and I look at what we're doing at Spotify, like it is truly amazing. I mean, the perfect music product is one that. I've described this many times, and I think we're getting closer to it now, which is awesome, Right now, you can put together a better playlist yourself than what the system could do. But I think in the future, downrightly, we can soundtrack every moment of your life in an amazing way where you just feel more. The power of great music is that you will feel more. If you're happy, you'll feel happier. If you're wallowing and feeling sad, you'll feel even more wallowing and sad in that moment. And that's thanks to AI. Like, you know, 950 moles, being able to track them over time, keeping you safe. Like, these are positive, amazing examples with AI. And yet we talk about this thing as we tend to amplify that. But I want to see more positive examples of AI. I think we as an industry have done a terrible disservice of not talking about all the really, really important. crazy positive stuff that we can use AI for that actually greatly benefits everyone. And in that sense, do you also think about the benefit and the value of open source AI, open weight models, Absolutely. I think you can now look at like a 13 cent per million token output cost versus $30. That's an incredible difference that unleashes this technology across a broad spectrum of businesses and individuals that there is no... concentration of value or wealth amongst a handful of businesses or individuals that if we can really make [40:00] these open weights and open source models proliferate. But obviously this regulatory conversation may restrict the expansion of open source and open weight models and if you have a view on kind of the importance there i think technology has always gone between open and closed right like we had windows versus linux in the first iteration you know um ios versus android like we've seen this play out so many times and where we tend to net out is we tend to have both um and you know to models before and I will keep on doing that because I think that that brings a lot of innovation that will be really helpful for the ecosystem as well. So I completely agree with you, Dave, that I think that if you look at these now and certainly what we're seeing You know, let's take Spotify as an example where we're obviously using a lot of the frontier models, but we have a bunch of fine tuned models ourselves, too. And we tend to do both. And that is actually making us more innovative. And that's actually allowing us to do certain things that we couldn't do with the frontier models, either because of cost and efficiency reasons or sometimes because we just couldn't tune in the way we would want. I personally think it's like, It's going to be both. But one of the things I heard that is a perspective I haven't heard before that maybe I would encourage people. Not that I am sort of an expert on all these things, but I'm surprised that people haven't talked about sort of the amount of computes. So if you think about it, for me, this isn't about just the intelligence itself of a single model, but it's still, you know, of the basis of... you know, the amount of computes you're doing can indicate something, right? Like, so if I'm using 100,000 GPUs for something, that's probably going to be a lot more powerful than if I'm running an open source model on my home PC. It's probably very unlikely that I'm going to be able to do an enormous cyber thing, certainly if the companies have mythos class models to try to, be able to with an open source model on my home pc do a lot of damage so you know i i just find it to be a fascinating thing to just think about if there's a way of um of looking at the amount of computes as well as one sort of factor in that. Right. Assuming model equivalency compute... As a key metric of defensibility, and that's really where we can start to build guardrails around who can have access to the most compute and how do you get certified for compute as opposed to certified for software. Wasn't that the original, like two years ago, one of the original concepts of limiting this was how many teraflops or how many GPUs you had? That was how they wrote the regulatory laws around regulating models. model. that was trained on X number of teraflops. And it was like, this was like a California-- - As if this is static. - Yeah. - Like nothing's gonna change. - California's genius assembly got together and said, "We figured it out. "We know how to classify scary models and bad models." - How many ethernet cables are in your data center as a proxy for how far it was. - How many watts of power went into making this model? But I do think your point of compute as a roadblock is probably a good one. This is what it used to be. You remember when Cray supercomputers came out? And there was a limited access and they were considered like a security risk and you have to get approved in order to get access to a Cray supercomputer. Because you could use them to design nuclear weapons and you could use them to crack codes and do all the things that people are worried about. Yeah, today a crazy supercomputer I think is less power than a washing machine or something, some product we probably have in our kitchen. It is a really great point actually, Daniel, because if you've got strong compute – installed as a defense capability, assuming model equivalency, you're going to be hugely advantaged. I haven't seen it in the debate, at least, and I'm surprised, but maybe it's something I'm missing. Hey, before we let you go, I guess two questions for you. Take them whichever order you want. What's the strategy and how does Spotify look at something we care about deeply, podcasting? which is built on open standards. And you guys have become a major player in it. open standards are kind of moving one direction and Spotify has got a lot of proprietary stuff. I've talked to you about this and some of your team members like, hey, can we keep supporting the open source port? And then the second piece, People don't know this, but there was this incredible company, Stardoll. It was one of the first companies in Stockholm, Sweden, to ever get funding from American VCs. Yeah. [45:00] CTO, if I remember, or the intern for a couple years? Well, the intern, CTO, whatever. It was probably one of those. People don't know this was an incredibly innovative company in terms of, you know, many variables like digital goods and subscriptions, all that stuff. So maybe take us through either both of those questions, because that's just, you know, two personal things that I would love to hear about if we were at dinner. yeah i mean look uh again as i said i think the the open and closed uh are going to be two things that are going to exist um side by side. So the interesting thing about Spotify today is like on the one hand, of course, we're a huge platform in podcasting and we have a destination. There are certain things that we're doing that general podcasts may or may not be doing. You know, one of those will be how we're handling comments and so on that are specific to the platform. But we actually do have tools. So what a lot of people don't know is that we also through our platform, aggregate and distribute podcast onto a lot of the other podcasting platforms so there's a bunch of podcasters they're using our tools to then distribute to some of the other platforms too and we're obviously doing that through open standards so i don't think the answer for us is like um one way or the other we think both can coexist but obviously if you're if you want to innovate within a standard then the key is to have other people agree to that standard and that can sometimes take you longer So Spotify's approach has been let's support both in tandem. So that's that. Then there's two features. I just wanted to make sure I ask you about. There's one for going live. So this is part of the new standard. So having that on our page, so when we go live, it actually respects the RSS feed and sends that note out. And then there's another one called value and donations where the person who has like, you know, if they happen to be doing... donations on whatever platform or they do them to paypal those two the like podcast ogs in the underground really want spotify to support so i'm just doing my job and representing as a 17 year podcaster please add those two or support those two if you think it's worthy of doing all right well it's no longer my decision alone i should say but i will certainly bring right to the chairman while i got him on the line yeah of course of course but yeah that was about starting out yeah I'll definitely bring the feature request to the team. Tell us about Stardall while you wrap up here. What did you learn there when you were 20 years old, 21, 22, whatever it was at Stardall? Yeah, you know, it was a great experience. It's kind of a, you know, it brings me back memory lane. You know, I I did start all just before I started Spotify. Um, so there was a entrepreneur called Matthias who, uh, was kind of one of the OG entrepreneurs, um, in the Swedish ecosystem. He had been around since the dot com bubble and he had, um, it was originally the site called paper doll heaven in Turco, Finland, uh, out of all places by this old lady and her son. And, um, Index Ventures had invested in it and taking kind of a majority stake to help doing this because they had no idea of scaling this site. And actually, you know, And Index and Matias kind of came in with the idea of rebranding it and building it into Startle. And so they came to me one day and they sort of asked around for people who knew tech and could be technical enough to help build it. And I had already decided I wanted to go build Spotify, but Danny Reimer is very good at convincing people to do things. By the way, it didn't end up happening. But... And, you know, I kind of said, okay, fine, I'll help you for, I think it was six to 12 months, something like that. So I went over to Finland and I checked it out. And it was kind of wild and crazy because I think the average page load time at that time was like four minutes. So it took four minutes to render a simple page. So it was basically the server was completely breaking down by its seams. I was like, holy shit, this is crazy. they made this work in a snappy way. And so we basically re-architected the website. I hired an entirely new technical team and brought it down to under a second of load time. And obviously the traffic exploded. Sequoia invested. And I said, okay, well, I've done my part. Thank you. And then I went off and built Spotify. It's amazing, these origin stories. Like what happened right before the thing is always interesting. [50:00] Listen, Daniel, continued success. If people want to try Neko, N-E-K-O-N-E-K-O, Where can they go? So we are launched the first one at 300 Lafayette. So go sign up on Nico health.com. Um, waitlist and we'll try to get to you as soon as possible. And this is in New York, 300 Lafayette Street. So only in New York so far, but we're expanding quickly both the number of venues we have in New York. We're launching in Florida as well, in Miami, and we're launching in D.C., and we're going to try to launch all over the U.S. in the coming 12 to 24 months. Yeah, it's great. Right on Houston Street. I'd say back in the day when Sudo had their big studios, they were right around the corner from you. I want to talk about memory lane. All right, continued success. Everybody go to NicoHealth.com and sign up for the waitlist. I'll see you next time. Bye-bye.