← Back to list
Colossus (Invest Like the Best / Business Breakdowns)Podcast21 Apr 2026Source: colossus.comHost: Patrick O'Shaughnessy

Alex Karnal - The Trillion-Dollar Health Revolution - [Invest Like the Best, EP.468]

In plain words

This interview says many deadly diseases already have treatments, but people need to use them proactively. Guest Alex Karnal argues GLP-1 drugs (like Novo Nordisk and Eli Lilly's) first proved people will pay to manage health—oral versions hit $150/month and sales soared. But he's more excited about PCSK9 inhibitors (cholesterol-lowering drugs), calling them 'the closest thing to a free lunch' with fewer side effects and bigger long-term potential than GLP-1s. He also notes AI drug-discovery firms (like Lila Sciences) can find drug candidates in a month, but only if their data is reliable.

AI SummaryAI-generated · may contain errors · verify against the original

At a Glance

Alex Karnal (co-founder of Braidwell) introduced the "health stack" framework on the program, arguing that most fatal diseases can already be addressed with existing drugs. The core thesis is: GLP-1 drugs are the first commercial proof that people are willing to proactively manage their health, while PCSK9 inhibitors, despite lower attention, may be an even more important drug class. He warned that most published literature used to train AI companies is irreproducible, but believes biology is moving toward deterministic scientific superintelligence. Key conclusions include: defensive health strategies are divided into five layers, GLP-1 represents a turning point, PCSK9 inhibitors are seen as the "closest thing to a free lunch" in medicine, and automated laboratories will drive a revolution in drug discovery.

~11 min full read · 6 sections
Deep Analysis

Theme 1: Health Stack – A Five-Layer Defense Strategy, Existing Drugs Already Address Most Fatal Diseases

Alex Karnal argues that drugs are already available for most diseases that cause human death, and the key lies in shifting from "passive treatment" to "active defense." He proposes the concept of a "health stack," comprising five core defense layers: lipid optimization, cardiometabolic health, neurocognitive health, inflammatory health, and blood pressure. Each layer has approved drugs, but the public has not fully utilized them.

  • Lipid Optimization: LDL cholesterol is a "silent killer," slowly accumulating in blood vessels and leading to heart attacks or strokes. The probability of a cardiovascular event in people aged 40–80 is 30%–50%, but existing drugs (statins, PCSK9 inhibitors) can reduce the risk to below 10%.
  • Cardiometabolic Health: A high-glucose environment makes blood vessels "brittle," creating a "multiplier effect" with lipid accumulation. GLP-1 drugs (e.g., from Eli Lilly) show data indicating a 94% reduction in the risk of progression from prediabetes to diabetes, and help reduce weight from obesity to normal levels.
  • Neurocognitive Health: Anti-amyloid drugs (e.g., upcoming data from Eli Lilly) are expected to clear brain plaques early and slow the progression of Alzheimer's disease. Karnal speculates that early intervention may show "dramatic" protective effects.
  • Inflammatory Health: Overeating (e.g., highly processed foods) drives systemic inflammation, not only exacerbating cardiovascular risk but also linking to inflammatory diseases such as atopic dermatitis and ulcerative colitis.
  • Blood Pressure: As the fifth layer, controlling blood pressure is a key component of the overall defense.

> Karnal's assessment: "Most diseases that will take our lives already have drugs that can address them. The gap is not in needing more drugs, but in how to direct these drugs toward the impact they can have."


Theme 2: GLP-1 — Commercial Proof That People Willing to Proactively Manage Health, Oral Version Ignites Demand

Karnal believes that GLP-1 drugs represent the "first commercial proof" that people are willing to take proactive action for long-term health. Three major findings in 2025 support this judgment: consumers prioritize "tolerability" over maximum weight loss, direct-to-consumer channels (such as Lilly Direct) contribute over half of new users, and price elasticity is enormous.

  • Market data: In 2025, the GLP-1 market adds 200,000 new prescriptions weekly; the oral version of Wegovy (launching in early 2026) pushes this figure to 300,000 per week, with monthly costs dropping to $150 (annualized at $1,800), causing sales to "fly off the shelves." In comparison, the early growth rate of the injectable Zepbound was only about one-quarter of that of the oral version.
  • Mechanism: GLP-1 is a natural human hormone, but its half-life is only 2 minutes. The scientific breakthrough lies in modifying it into a once-weekly injection or once-daily oral pill. It acts on the stomach (slowing digestion) and the brain (inducing satiety), while also regulating insulin secretion and protecting the kidneys and blood vessels.
  • Market segmentation: Karnal divides the obesity market into three segments — overweight (BMI 25-30), obese (30-40), and morbidly obese (>40). Existing drugs "serve well" the first two segments, but the morbidly obese require higher doses and longer titration times, which is where the "right to win" lies.
  • Benefits independent of weight loss: Novo Nordisk data shows that GLP-1 reduces the risk of heart attack/stroke by over 20%, and this effect is independent of weight loss. Additionally, protective effects against addiction (alcohol, drugs, gambling) are emerging in the data.

> Karnal's assessment: "People are not looking for the most powerful 'rocket launcher'; they are looking for something that allows them to lose weight, stabilize, and stick with it long-term. Tolerability is more important than maximum weight loss."


Theme 3: PCSK9 Inhibitors—The Closest Thing to a "Free Lunch" Drug, with Potential Surpassing GLP-1

Karnal believes that PCSK9 inhibitors are the "closest thing to a free lunch" in pharmaceuticals, and in the long term, their market may be larger than that of GLP-1s, though current adoption is slow due to "invisible benefits." By lowering LDL cholesterol by 50%, this drug reduces the future risk of heart attack or stroke by over 20% in patients with a history of such events, and by approximately 25% in high-risk populations.

  • Discovery Pathway: Rooted in human genetics—individuals carrying mutations in the PCSK9 gene (who do not produce the PCSK9 protein) showed a 88% reduction in cardiovascular disease risk in a 15-year longitudinal study. The drug aims to replicate this natural advantage.
  • Technological Evolution: From injectable monoclonal antibodies (requiring 26 doses per year) to RNA interference technology (requiring only 2 doses per year), significantly improving compliance.
  • Why a "Free Lunch": Animal models show that LDL can be reduced to near zero without complete disappearance; individuals with genetic deficiencies remain healthy long-term. Karnal argues that the side effect risk of PCSK9 inhibitors is far lower than that of GLP-1s (the latter being associated with nausea, vomiting, diarrhea, gallstones, pancreatitis, and muscle loss).
  • Adoption Barriers: High cholesterol is a "silent killer"—patients feel no symptoms, whereas GLP-1s provide immediate weight loss feedback. Karnal asserts: "If you look only at the long term, PCSK9 inhibitors should surpass GLP-1s in both patient numbers and revenue."

> Karnal asserts: "PCSK9 inhibitors are very close to a free lunch. GLP-1s have real toxicity, but for PCSK9—we have human genetic evidence that people who don't produce this protein live longer and healthier lives."


Theme 4: AI and Automated Laboratories – Drug Discovery Is Heading Toward "Scientific Superintelligence"

Karnal believes that the greatest value of AI in drug discovery lies in breaking the constraints of "the human brain and hands," but this requires high-quality, reproducible data. He warns that most published literature is "irreproducible," so AI companies relying solely on public data will fall into a "garbage in, garbage out" predicament.

  • Current Progress: AI can now achieve "from model to molecule"—compressing the traditional screening process that takes years into one month. Representative companies such as Lila Sciences and Anabla can rapidly generate candidate molecules for known targets.
  • Key Bottleneck: Training AI requires "scientific tokens" (high-quality experimental data), yet a large number of experiments in public literature are irreproducible. Karnal emphasizes that successful AI companies must have proprietary ways to "generate scientific tokens," which will become their moat.
  • Future Vision (5–10 years): Fully automated laboratories—robots running experiments 24/7, with AI agents driving hypothesis generation and validation. Scientists will be freed from "the pipette" and focus on "pointing in the right direction for breakthroughs." Karnal argues that this path follows a "deterministic scaling law," making scientific superintelligence "almost inevitable."
  • FDA's Alignment: Commissioner Marty Makary is pushing reforms—using AI to accelerate reviews, reduce animal testing requirements, and consolidate redundant studies. Karnal believes this reflects a trend where "the entire system is pursuing speed and efficiency."

> Karnal's assessment: "We are on a deterministic curve toward scientific superintelligence in biology. AI means we are no longer limited by brain and hands—we can generate all possible hypotheses and run experiments at an unprecedented scale."


Mentioned Positions

Position Guest Stance Key Data
Eli Lilly Bullish (GLP-1 leader) Oral Wegovy monthly cost $150, weekly new prescriptions increased from 200k to 300k; risk of progression from prediabetes to diabetes reduced by 94%
Novo Nordisk Bullish (GLP-1 pioneer) Heart attack/stroke risk reduced >20% (independent of weight loss)
Amgen Neutral (under watch) Developing GLP-1 dosed monthly/quarterly
Pfizer Neutral (under watch) Acquired a monthly GLP-1 candidate
Biogen Neutral (Alzheimer's field) Cognitive decline slowed by ~30% in late-stage patients
Exact Sciences Bullish (colorectal cancer screening) Stool test has high sensitivity, but completion rate limited by sample mailing
Guardant Health Bullish (colorectal cancer screening) Blood test "is accelerating," addressing Exact Sciences' completion rate issue
Capstan Bullish (next-generation CAR-T) No cell extraction needed; intravenous injection attacks cancer cells in vivo; tumor shrinkage of 100% achieved in >70% of patients
Lila Sciences Bullish (AI drug discovery) From model to molecule within one month
Anabla Bullish (AI drug discovery) Same as above
Neko Neutral (skin imaging) No specific data provided
Prenuvo Neutral (whole-body imaging) No specific data provided; Karnal personally has not adopted it, concerned about false positive rate

Judgments Worth Remembering

1. The “Health Stack” Five-Layer Defense (Karnal): Lipid optimization, cardiometabolic health, neurocognition, inflammation, and blood pressure—each layer has approved drugs that are underutilized by the public. Support: For individuals aged 40–80, the 30%–50% risk of cardiovascular events can be reduced to below 10% with existing medications.

2. GLP-1 Is the “First Commercial Proof” (Karnal): People are willing to take proactive action for long-term health. Support: When the monthly cost of oral Wegovy was $150, weekly new prescriptions surged from 200,000 to 300,000; compounded GLP-1 accounts for 15%–20% of the market, demonstrating significant price elasticity.

3. PCSK9 Inhibitors Are the “Closest Thing to a Free Lunch in Drugs” (Karnal): A 15-year follow-up of individuals with genetic deficiencies showed an 88% reduction in cardiovascular risk, with no known serious side effects. Support: Animal models indicate that LDL does not drop to zero; the long-term patient base and revenue potential for this drug should exceed that of GLP-1.

4. The Benefits of GLP-1 Are Independent of Weight Loss (Karnal): Novo Nordisk data show a >20% reduction in heart attack/stroke risk, unrelated to weight loss. Support: Protective effects against addiction (alcohol, drugs, gambling) are emerging in the data, which cannot be achieved through dieting alone.

5. The Key Bottleneck in AI Drug Discovery Is Data Quality (Karnal): Most published literature is not reproducible; AI companies relying solely on public data will fall into a “garbage in, garbage out” trap. Support: Successful AI companies must have proprietary methods for generating “scientific tokens,” which form a competitive moat.

6. Automated Labs Will Achieve “Scientific Superintelligence” (Karnal): Within 5–10 years, AI agents combined with robotics will run experiments 24/7, freeing scientists from pipetting. Support: Screening that traditionally took years has been compressed to one month; this path follows a “deterministic scaling law.”

7. The “Sensitivity vs. Specificity” Trade-off in Cancer Screening (Karnal): Multi-cancer early detection is advancing, but the false positive rate remains a major risk. Support: Colonoscopy is still the gold standard; blood tests (e.g., Guardant) are accelerating, but Karnal personally avoids whole-body imaging (e.g., Prenuvo) due to concerns over false positive interference.

8. “Citizen Pharmacology” Runs Parallel to FDA Reform (Karnal): Self-experimentation on platforms like Reddit reflects a trend of “people not wanting to wait,” while FDA Commissioner Marty Makary is pushing for accelerated reviews and reduced animal testing. Support: The FDA is introducing AI systems to review thousands of pages of filings; Karnal believes the entire system is “pursuing faster, more efficient” processes.