Jan 2, 2019, 12:41 AM ETPodcast, Interview

a16z Podcast | Taking the Pulse on Bio

A16Z is pivoting its bio investment thesis away from empirical science toward an engineering discipline, explicitly aiming to remove "science risk" (novel biology risk) in favor of managing engineering and scale risk.
Machine learning and AI are identified as primary tools to replace human hypothesis testing, allowing algorithms to learn underlying biological complexities that are currently beyond human comprehension.
The firm distinguishes between high-science-risk diseases (e.g., Alzheimer's, where mechanisms are unknown) and lower-science-risk opportunities (e.g., Sickle Cell Anemia, where the mutation is known and CRISPR/gene therapy offers a direct engineering solution).
Engineering approaches enable the generalization of drug development; unlike bespoke traditional therapeutics, engineering platforms (like CRISPR or mRNA) allow knowledge to transfer across targets by simply "swapping the code" rather than rediscovering biology for each new molecule.
The team expects the shift to transform biology from an unpredictable, creative science into a repeatable, schedulable engineering process, similar to how the semiconductor industry evolved.

AI applications in diagnostics are moving beyond "better or faster" to enabling previously impossible clinical insights, such as using genomics and immune system data to detect cancer early with higher accuracy than human physicians.
A notable trend is the convergence of new data sources (wearables, genomics, retina imaging, facial features) with AI to create actionable, scalable health solutions.
Companies like Freenome demonstrate the engineering model: once a genomic test is validated for one cancer type, the exact same engineering process can be repeated for other cancers without developing new biology, contrasting sharply with traditional diagnostics where science must be reinvented for each disease.
Data network effects are becoming a primary barrier to entry, where increased data volume directly improves algorithmic accuracy (e.g., from 85% to 97%), creating a defensible moat that persists even as patent windows expire.
Digital health therapeutics are utilizing A/B testing in "computer land" to iterate efficacy without toxicity, as seen in behavioral therapy platforms like Omada managing Type 2 diabetes.

Digital health startups typically follow a go-to-market progression: evidence generation $\rightarrow$ reimbursement $\rightarrow$ commercial traction, often starting with self-insured employers before moving to insurance plans and government payers (CMS/Medicare).
Self-insured employers initially served as early adopters because they value both cost savings and employee engagement, whereas traditional health plans prioritize cost reduction alone.
Network effects are being applied to healthcare infrastructure; for example, PatientPing connects providers to coordinate care, where the value of the network increases as more providers join, allowing granular patient tracking.
The "laboratories of democracy" model is observed in regulation, with state-level adoption often preceding federal (CMS/Medicare) integration.
Regulatory risk is increasingly viewed as a proxy for scientific risk; as therapies become more effective (e.g., CAR-T, gene therapies for blindness), FDA approval rates have become unanimous (13-0 votes), signaling that proven efficacy minimizes regulatory hurdles.

Engineered Cell Therapies: The emergence of "living drugs" that possess logic (targeting, acting, and self-terminating) represents a shift from passive biologics to intelligent therapies.
Bio-Circuits: The future involves designing biological circuits inside cells using Electronic Design Automation (EDA) flows, aiming to replicate the speed of innovation in electronics within biology.
Preclinical Models: A paradigm shift is occurring away from animal models toward "organs on chips" or "humans on chips" to create more predictive testing systems, which aims to reduce clinical trial failure rates.
Clinical Trial Efficiency: Technology (e.g., social networks for recruitment) can reduce patient recruitment costs and time by over 90%, compressing development cycles.
Non-Therapeutic CRISPR: CRISPR is expanding beyond direct disease cures into diagnostics (infection detection) and drug discovery platforms.
Interdisciplinary Founders: A16Z prioritizes founders with deep dual-experience in both computer science and biology, viewing this hybrid expertise as critical for success in the "bio-engineering" era.
Systemic Interoperability: A long-term vision includes fully interoperable Electronic Health Records (EHRs), such as Epic and Cerner systems communicating seamlessly.

A16Z is actively seeking entrepreneurs building companies where the fundamental biology is understood, allowing the venture to focus on engineering execution rather than de-risking novel science.
The firm anticipates a convergence of tech and biotech investing, where companies are valued on both technical scalability (market risk) and de-risked biological efficacy.
The team notes that innovation cycles in biology are accelerating dramatically, with more breakthroughs occurring in the last two years than in the previous decade, creating a "breathtaking" pace of new opportunities.
New modalities (engineered cells, mRNA, CRISPR) are expected to dominate, moving beyond small molecules and traditional proteins to create entirely new classes of therapeutic tools.