The Debate on AI and Cancer Cure

By some estimates, over a trillion dollars have already been invested in artificial intelligence. Despite this, large tech companies like Meta and OpenAI continue to pursue more powerful and versatile AI, aiming to match or exceed human performance. A considerable amount of resources is being poured into developing Artificial General Intelligence (AGI) or even Artificial Super Intelligence (ASI, fueling high expectations for their potential capabilities.

Excitement around these technologies is often accompanied by bold claims, including the possibility of curing cancer. This specific scenario has caught the attention of Emilia Javorsky, director of the Futures program at the Future of Life Institute, a think tank focused on the benefits and risks of transformative technologies like AI. Her perspective offers a critical yet optimistic analysis of AI's application in medicine.

Understanding “Curing Cancer” with AI

In March, Javorsky published an essay titled “AI vs Cancer,” drawing on her experience as a doctor, scientist, and entrepreneur. The text critiques the excessive faith and resource allocation into artificial super intelligence as a future solution for diseases. Javorsky highlights how numerous factors, beyond mere computational intelligence, limit the development of new treatments and access to innovative care. AI, in fact, cannot analyze patient data that was never collected, and any treatment is flawed if patients risk bankruptcy seeking it. However, the essay also aims to instill optimism by showing how existing forms of AI are already being successfully applied in the fight against cancer.

Javorsky emphasizes that cancer is not one universal disease curable by a single treatment. Instead, it is a highly individualized co-evolutionary process, where different mutations drive the pathology in each person and even within a single tumor. Solutions, therefore, will need to be personalized. In medicine, we have not yet “cured” complex chronic diseases like diabetes or heart disease, but we have developed effective ways to manage them. The medical community's hope is to turn cancer into a chronically manageable condition, no longer a death sentence, through personalized and effective treatments. It is crucial to distinguish between the promises of a future “super-intelligence” and the concrete capabilities of today's AI, which is already accelerating innovation in drug discovery, drug toxicity prediction, biomarker definition, clinical trial acceleration, and early detection. These tangible advancements risk being overshadowed by unrealistic expectations related to AGI or ASI.

Investing Pragmatically: Data and Infrastructure

In a world with finite capital, allocating resources to curing cancer represents one of the noblest goals. Javorsky argues that it is crucial to identify where to invest to get the maximum return on investment and effectively solve the problem. In her view, there is an overinvestment in the “intelligence-compute” side of things and an underinvestment in innovating tools to measure biology and in creating high-quality, large-scale datasets. The current healthcare system is fundamentally a “sick care” system, where data is collected only when people become ill. This approach limits AI's ability to operate at its full potential.

For those evaluating on-premise deployment of AI solutions, the issue of collecting and managing high-quality datasets is central. Data sovereignty and the ability to process large volumes of sensitive information in controlled environments become critical factors. Investing in robust local infrastructures, capable of supporting the acquisition and analysis of complex biological data, can represent a competitive and strategic advantage. The need for a holistic approach to medicine, including broader and more proactive data collection, is a prerequisite for unlocking AI's true potential. In an ideal world, all paths could be pursued, but reality dictates strategic choices in capital allocation. Javorsky expresses confidence in AI, but only if investments are directed towards the right types of AI and the true bottlenecks of the system.

A Roadmap for the Future: Optimism and Action

Javorsky concludes her essay by proposing a clear roadmap for addressing the cancer problem. Her vision is structured around three fundamental pillars. The first involves resourcing and scaling the AI tools that are already demonstrating significant progress in oncology. This includes AI for early detection, accelerating clinical trials, and developing in silico models and “digital twins” for personalized medicine, which aim to create high-fidelity digital representations of patients to optimize individual treatments.

The second pillar calls for doubling down on investments in promising areas of biology related to oncology. Finally, the third point focuses on tackling the institutional and systemic bottlenecks that hinder medical progress. This perspective, while acknowledging the challenges, aims to instill a sense of optimism, highlighting that reality already offers many reasons to hope for a future where AI, if strategically employed, can contribute decisively to cancer management and treatment.