Breyer Capital’s Healthcare AI Investment Thesis: Learnings and Predictions (Fall 2023 Update):
Since our last Healthcare AI investment thesis update, we’ve experienced a continuing and profound transformation at the intersection of healthcare, life sciences, and artificial intelligence (AI). At Breyer Capital, we’ve been investing in companies and teams operating at this intersection since 2016. This past spring, our work helped catalyze the inaugural University of Texas-Breyer Capital-UTIMCO Healthcare Summit, a two-day event in Austin, Texas featuring perspectives from medical experts, top AI researchers, and healthcare and life sciences executives. Presentation topics included CAR-T and gene editing, AI in drug discovery and development, applications of clinical AI, quantum technologies for life sciences, ecosystem-building in biomedicine, and much more. We are deeply appreciative of our partners and colleagues who contributed to making the event a remarkable success. We look forward to welcoming folks back to Austin in Spring 2024!
In our last post, we offered an updated view of our investment thesis, integrating learnings and best practices inspired by growing our portfolio. We highlighted emerging interest in multi-modal healthcare AI models — those that combine multiple types of data spanning clinical, -omics, and imaging, to name a few — and discussed how AI can facilitate a more personalized, empathetic care delivery system.
Underscoring remarks from our last update, we remain extraordinarily optimistic that the application of computation in healthcare and life sciences will be the financial and impact opportunity of this decade, and now likely, the next.
We are very excited about the ways that the launch of ChatGPT4 by OpenAI has accelerated the understanding of AI and catalyzed enthusiasm for adoption across industries. And, as generative AI has swept headlines, the interdisciplinary innovation in healthcare and life sciences has captured the imagination of our country’s most talented AI research scientists and industry lab leaders, many of whom are making long-term bets on healthcare.
We are also optimistic that the field of oncology will continue to shine light on the path toward molecular medicine. In tandem, we expect the fields of cardiology and neuroscience to make immense strides towards precision medicine. Within neuroscience, we predict that our rapidly maturing diagnostic and therapeutic toolkits will reveal opportunities to treat even our most intractable diseases, especially those in cancer, cardiology, and the neurodegeneration space such as Alzheimer’s Disease and dementia.
There’s no question that the COVID-19 pandemic exposed critical gaps in our nation’s healthcare infrastructure, revealing structural and systemic shortcomings that will require highly scalable solutions to address effectively. Very different from our discussions in 2016 and 2017, as we spend time with physician leaders and top hospital executives, we’re now noticing an unparalleled receptivity to new ideas in healthcare. Notably, there is excitement and optimism about how AI can transform the medical field by addressing pertinent challenges such as clinician burnout, hospital operations, patient access, and revenue cycle management.
During the pandemic, we also observed how AI can accelerate innovation in the life sciences. Moderna leveraged AI across discovery and development workflows for the COVID-19 vaccine spanning mRNA optimization, rapid prototyping, preclinical testing, manufacturing, and real-world monitoring tasks, ultimately predicting the protein structure of the SARS-CoV-2 spike protein using computational methods. Though AI-driven discovery is a concept that dates back to the 1980s (previously referred to as “Computer-Aided Drug Design”), Moderna demonstrated on a global stage both the power and potential of AI-driven discovery and development in life sciences, a category now often referred to as “AI-first biotech.” Already, over the last decade, the initial cohort of AI-first biotech companies has generated over 70 AI-derived clinical pipeline assets spanning small molecules, antibodies, and vaccines. Since 2019, AI-first biotech companies have garnered over $10 billion in venture investment, representing only 1/10th of the total capital deployed at the intersection of biomedicine and AI.
From our perspective, we’re at the earliest innings of the AI paradigm shift in drug discovery.
Breyer Capital remains grateful to be working closely with technology leaders such as Marc Benioff, Sergey Brin, Tim Cook, Michael Dell, Judy Faulkner, Jensen Huang, Arvind Krishna, Doug McMillon, Satya Nadella, Larry Page, Sundar Pichai, Sheryl Sandberg, Eric Schmidt, Stephen Schwarzman, Mark Zuckerberg, and so many others. Our colleagues at the helm of leading technology companies share our passion and belief in the opportunity in healthcare AI, evidenced by significant investments in industry-leading research efforts across large-cap tech companies. Recently, Microsoft Research launched LLaVA-Med, a multi-modal biomedical assistant for clinical text and imaging data. Alphabet has continued to strengthen MedPaLM, a large language model trained to answer medical questions, generate text about medical topics, and translate between medical languages. Collaboration is also flourishing, with OpenAI, Microsoft, and Epic partnering to bring generative AI capabilities to the electronic health record. Big tech’s embrace of healthcare and life sciences is a boon for this nascent category.
More than ever before, we believe that startup companies born out of collaborations among top academic researchers, scientists, and early-stage operators will bring forth some of the most influential and impactful advancements in AI and biomedicine.
Over the next year, as we have done in recent years, we expect to double and triple down on our investment thesis at this critical nexus partnering with excellent clinicians, scientists, AI researchers, startup operators, and AI safety and medical ethics experts.
Already, in addition to the “public” portfolio companies, we have a number of “stealth” investments that in many cases are spin-outs and spin-offs from our best medical schools, universities, and institutes.
As we head into the final months of 2023, we recommit to key focus areas for Breyer Capital:
- Drug Discovery: Discovering novel therapeutics using wet and dry lab techniques hand-in-hand, with the end goal of making medicines, and especially those addressing historically “undruggable” targets with significant unmet clinical need.
- Phenotyping: Developing AI models that are capable of characterizing disease with greater precision, such as knowledge graphs for precision medicine. These tools can unlock insights that generate value for therapeutic initiatives or as novel biomarkers or clinical diagnostics.
- Clinical Trials: Authoring AI models that synthesize optimized protocols, structure clinical data to streamline trial recruitment (including novel AI biomarkers), and facilitate real-world evidence studies and virtual control arms.
- Back Office Automation: Unburdening the healthcare system from the ~$1 trillion of administrative waste via AI models that tackle fraud and waste, back office inefficiencies, and fax-driven data exchange.
- Supercharging Clinicians and Scientists: Equipping front-line workers in healthcare and life sciences with AI-infused tools that abstract away administrivia, allowing them to focus on what they trained to do: care for patients and perform great scientific research.
We also remain open-minded to founders and researchers showing us new areas for consideration.
Breyer Capital is appreciative of our enduring collaborations with superb investors and technology and healthcare leaders such as Tom Cahill, Gerald Chan, Marty Chavez, Morgan Cheatham, Helena Foulkes, Alan Hutchison, Steve Pagliuca, and Sam Palmisano, among others. We are also humbled by the vision for biomedicine championed by our many academic collaborators including Fei Fei Li, George Daley, Zak Kohane, and many others.
Over the next few months, we will continue to exchange ideas and investment perspectives with leaders at some of the most forward-thinking tech companies, brainstorming and thinking through long-term, generational opportunities in AI.
We’d be remiss not to mention our enthusiasm for the remarkable work coming out of the Austin biomedical, entrepreneurship, and investing ecosystems following our Spring 2023 event and most recently evidenced by the recent announcement of The University of Texas at Austin Medical Center, a collaboration between the University of Texas at Austin and MD Anderson Cancer Center to build two new hospitals in downtown Austin that will “undoubtedly result in transformative cancer care.” It is a truly remarkable life-sciences and computation collaboration.
What began as an emerging investment thesis for Breyer Capital in 2016 has blossomed into a pillar and area of expertise for our firm, spanning early-stage venture capital investing to large-scale philanthropic initiatives. In closing, the future of biomedicine couldn’t be brighter.
But what might it actually look like?
We asked some of the best academic and industry thought leaders and some of our leading healthcare AI portfolio companies — here’s what many of the CEOs told us:
- Artera: “Artera envisions a future of biomedicine where AI-empowered diagnostic tests transform healthcare by improving patient outcomes and personalizing therapies. We foresee a revolution in histopathology, where our multimodal computer vision models analyze digitized images to predict disease trajectory and therapeutic response with unprecedented accuracy. Ultimately, we aspire towards a medical system in which every medical decision is informed by precise, AI-enhanced insights, making healthcare more effective, personalized, and proactive.”
- Atropos Health: “The future of medicine will be driven by personalized evidence for care. Individualized treatment plans will be driven by rapid evidence generation based on millions of similar patients. Generative AI will play a critical role in ensuring user experience helps uplift the user experience, removing the technology burden from physicians delivering care.”
- Earli: “What if we could stop the search for early cancer, relying on — often elusive — natural biomarkers? What if instead, we could *force* the cancer to reveal and ultimately destroy itself? A new generation of molecular biology, engineering, and software makes it possible for us to *use* the disease to turn it against itself, and only against itself.”
- Glass Health: “We envision a future where every clinician is empowered with AI clinical decision support that improves patient outcomes and decreases provider burnout worldwide.”
- Iterative Health: We envision a future where patients receive high-quality care no matter who they are or where they live. Technology and science will advance diagnostics to enable more precise pairing of patients to appropriate therapeutics to mitigate toxicities and accelerate quality of life.
- Paige: “Paige envisions a future where information about every single person’s cancer is available at the point of diagnosis by using AI, including grading, prognosis, and precision medicine treatment recommendations. AI can provide answers about biomarker status, prognostic markers, and predictive relationships that a pathologist otherwise would not be able to see. By utilizing tissue-based AI to query the essence of the tumor, we can accelerate time to treatment, improve outcomes, reduce costs and unnecessary treatments, and ultimately save lives.”
- Sandbox AI: “As Quantum computers start to get more fault-tolerant, bigger, and more robust, we’ll start to see a major impact on healthcare and life sciences…I think we’re in for a wonderful decade of innovation. Quantum computers still need to get to base-level maturity, but starting in about five years, they will start to really rock and roll. They will complement the tools we have today. We’re not waiting for quantum computers — right now, we are doing life-sciences, drug development, and battery chemistry work on Nvidia GPUs. That’s a very exciting development.”
- Soley: “Soley combines the biology of cellular sensing and anticipation, with AI-tech capabilities. This unique combination is used to de-code the cellular language that describes mechanisms of action and enables the forecasting of molecules that have the highest possibility to become successful drugs.”
- Subtle Medical: “At Subtle Medical, we envision a future where biomedicine is revolutionized by AI, with innovative solutions driving optimization in radiology workflows and reforming imaging pharmaceutical applications. AI technology such as Subtle’s AI solutions, can significantly enhance the efficiency, improve data quality and the precision of diagnostic procedures, reducing the necessary dosage of radiology drugs and thus minimizing potential side effects. In our vision, the latest technology developments will democratize access to advanced healthcare, ensuring all patients, regardless of location or socio-economic status, can benefit from high-quality diagnostics and treatments.”
- Xyla: “Information overload is a prevalent challenge in many high-value domains. It is an extreme challenge in medicine. In the year 2000, just shy of 2,300 clinical trials were conducted; by 2022, that number increased more than 10x, with 32,000 clinical trials started in 2022 alone. The amount of medical research published annually is doubling every 5 years. Such a rate of change makes it nearly impossible for the world’s physicians, medical scientists, and healthcare professionals to monitor and understand all the latest research literature and clinical evidence relevant to their work. OpenEvidence is our contribution to the global effort of taming the medical information firehose: Our approach uses artificial intelligence to aggregate, synthesize, and visualize clinical evidence in understandable, clinically useful formats that can be used to make more evidenced-based decisions and improve patient outcomes.”
We could not be more excited about another several years of partnering with companies leveraging computation from bench to bedside.
Jim