Table of Contents
For decades, the field of Alzheimer's research has been a graveyard of capital and ambition. Tens of billions of dollars have been poured into clinical trials, yet we still lack a cure. The prevailing narrative suggests that the disease is intractable, a biological puzzle too complex to solve. However, a new wave of thinking suggests that the lack of progress isn't due to a lack of clues, but rather a failure to connect them.
Sacha Schermerhorn, founder and CEO of Babylon Bio, argues that the answers we seek may already exist within the fractured landscape of scientific literature. By combining a fresh biological perspective with the synthesis capabilities of large language models (LLMs), Babylon Bio is attempting to rewrite the playbook on neurodegenerative disease. Their approach challenges the established dogma of the last century, proposing that we haven't failed because the problem is impossible, but because we have been trying to put out the wrong fire.
Key Takeaways
- The Amyloid Misconception: While amyloid plaques are a hallmark of Alzheimer's, targeting them is akin to blowing out a match after the house is already on fire. The focus must shift to Tau pathology to stop cognitive decline.
- Data Hidden in Plain Sight: Critical clues, such as the correlation between the shingles vaccine and reduced dementia risk, exist in siloed scientific fields. AI can now bridge these gaps.
- Missionaries Over Mercenaries: solving "impossible" problems requires a team motivated by scientific curiosity and impact, not just financial compensation.
- Commercial Intermediates: To fund a moonshot without diluting the mission, biotech companies can leverage "commercial intermediates"—repurposing existing drugs to generate revenue that funds long-term research.
- Digital Minimalism: Preserving "life minutes" by eliminating distractions, such as smartphones, allows for the deep focus required to tackle existential problems.
Rethinking the Alzheimer’s Dogma
Since Dr. Alois Alzheimer first characterized the disease in 1906, the field has been obsessed with two specific pathologies: amyloid plaques and neurofibrillary tangles. For the better part of the last few decades, the "Amyloid Cascade" hypothesis has dominated research. The theory posits that the accumulation of beta-amyloid is the primary driver of the disease, and removing it should cure the patient.
However, reality has proven far more stubborn. Drugs like Aducanumab have successfully cleared amyloid from the brain by significant margins, yet patients showed virtually no improvement in cognition. Schermerhorn offers a compelling analogy to explain this disconnect.
"If you were to go into a burning house, it doesn't matter what started the fire. It doesn't matter if it was a match or a toaster... it's on fire so you should put it out. The field has not updated their priors that the cause is actually not the thing that you need to target."
The Pivot to Tau
While amyloid may initiate the process decades before symptoms appear, it is the Tau proteins that correlate with the actual burning of the house—the cognitive decline. In a healthy brain, Tau helps stabilize the internal structure of neurons. In Alzheimer's, these proteins become hyperphosphorylated, detaching and forming tangles that choke the cell.
Schermerhorn hypothesizes that these Tau fibrils act like sponges, sequestering essential proteins that neurons need to survive. This structural collapse, rather than the initial amyloid trigger, is what destroys memory and cognition. By shifting the target from the match (amyloid) to the fire (Tau), Babylon Bio aims to intervene during the clinical phase where it matters most to patients.
Connecting the Dots with Artificial Intelligence
One of the profound inefficiencies in modern science is the sheer volume of data. Thousands of papers are published weekly, making it impossible for any single human expert to synthesize all available knowledge. This leads to "orphaned" clues—vital pieces of information that remain unconnected because they reside in disparate fields.
A prime example is the data regarding the shingles vaccine. Epidemiological studies in Wales showed that people who received the vaccine were roughly 20% less likely to develop dementia than those who missed the eligibility cutoff by just a week. This suggests a link between viral infection, the immune system, and neurodegeneration.
The Role of LLMs in Discovery
This is where Large Language Models (LLMs) become a lever for discovery. Schermerhorn envisions using AI not just to generate new biology, but to perform "Swanson Linking"—finding logical connections between unconnected nodes in the scientific graph. If Premise A is connected to Premise B in immunology, and Premise B is connected to Premise C in neurology, an AI can infer a connection between A and C that a human specialist might miss.
"We have more than enough clues to figure out the perfect target for something like Alzheimer's... Often these clues come from very disparate fields that don't interface with each other directly."
By training models on clinical trial readouts—the closest thing biology has to a binary "pass/fail" dataset—researchers can fine-tune AI to predict which biological mechanisms are most likely to succeed in a clinical setting.
The Economics of a Moonshot
Developing a new drug for a disease as complex as Alzheimer's is a multi-decade, billion-dollar gamble. Traditional biotech financing often forces founders to dilute their ownership or partner with pharmaceutical giants just inches from the finish line. To avoid this, Babylon Bio employs a strategy of "commercial intermediates."
The concept is similar to how Google Search (a cash cow) funds Google X (moonshots). In the biotech context, this means identifying existing drugs or "assets" sitting on shelves—compounds that have already cleared safety hurdles but were shelved for business reasons or lack of efficacy in their original indication.
By using their biological insights to repurpose these assets for new indications, the company can generate revenue and bolster its balance sheet. This financial independence allows the team to pursue the high-risk, high-reward Alzheimer's cure without being beholden to short-term investor demands or forced partnerships.
Building a Culture of "Missionaries"
Schermerhorn is candid about the difficulty of the early days—funding experiments with personal savings and facing rejection from countless investors. This adversity, however, became a powerful filter for talent. When a company cannot offer top-tier market salaries or guaranteed success, it attracts a specific type of person: the missionary.
The hiring philosophy at Babylon Bio centers on finding individuals who are obsessed with the problem itself. Schermerhorn looks for "drug hunters" who have failed at retirement multiple times simply because they love the science too much to stop. These are people who will exchange emails at 4:00 AM on a Saturday not because they have to, but because they are chasing a breakthrough.
"The best people in the world really are just so down to chat science. The money is not what drives them."
The Philosophy of "Life Minutes"
To maintain the intensity required for such a massive undertaking, Schermerhorn practices extreme digital minimalism, famously operating without a smartphone for the last three years. This decision is rooted in the concept of "life minutes"—the finite currency of our existence.
By eliminating the constant dopamine loops and distractions of modern technology, the team creates space for deep work and agency. Burnout, Schermerhorn argues, does not come from working hard; it comes from a lack of agency or the feeling of losing. When a team is deeply aligned with a mission and retains control over their work, their capacity for effort is nearly infinite.
Conclusion
The quest to cure Alzheimer's is not merely a scientific challenge; it is an information challenge and a structural one. The failure of the last thirty years was not a waste, but a necessary process of elimination that has highlighted where not to look. By pivoting focus to Tau pathology, utilizing AI to synthesize fractured human knowledge, and structuring a company to withstand the long timelines of drug development, Babylon Bio represents a new generation of biotech.
The clues are there. The science is tractable. It is now a matter of connecting the dots and having the endurance to see the mission through to the end.
