F Cancer

2025-06-07 · Source: Marcus on AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, AI in Medical Research · Depth: Intermediate, short

Summary

Gary Marcus critiques the current state of AI in cancer research, drawing heavily from an essay by physician-scientist Emilia Javorsky. Marcus highlights Javorsky's argument against "technosolutionism," which posits that AI alone can magically solve complex medical problems like cancer. The core issue, according to Javorsky, is not merely scientific but also involves systemic misunderstandings of healthcare, market failures, and economic incentives. Examples include the unprofitability of new antibiotics, the high cost of developing drugs for rare diseases, and patent expiration deadlines hindering investment in effective treatments like Tanespimycin. Marcus agrees with Javorsky's points on market forces and the naivete of oversimplifying biology, noting that 92% of animal drug trials fail in humans due to toxicity. However, Marcus also contends that Javorsky understates the need for more advanced AI capable of reasoning about chemistry, physics, and biology, beyond current large language models, to truly make breakthroughs.

Key takeaway

For AI Product Managers developing healthcare solutions, recognize that purely algorithmic approaches will not solve complex diseases like cancer. Your focus should shift from "magic algorithm" fantasies to addressing systemic issues, including market incentives and the profound complexity of human biology. Prioritize AI applications that enhance human researchers and streamline existing processes, rather than attempting to reinvent the entire medical system with a limited understanding of its underlying challenges.

Key insights

AI's impact on cancer is limited by technosolutionism, market failures, and insufficient biological understanding.

Principles

• Biology is inherently complex and highly individualized.
• Market incentives often hinder drug development.
• AI needs deeper reasoning for scientific discovery.

In practice

• Streamline administrative tasks in medical research with AI.
• Enhance human researchers with better AI tools.

Topics

• AI in Healthcare
• Cancer Research
• Drug Development
• Technosolutionism
• AI Limitations

Best for: AI Ethicist, AI Product Manager, Research Scientist

Related on AIssential

• He Raised $70M to Cure Every Disease With AI — AI agents, particularly specialized models, significantly accelerate scientific discovery by removing human talent bottlenecks and enabling high-throughput reasoning.
• Survey reveals concerns around AI impacting role of doctors - Irish Medical Times — Public support for AI in healthcare is high for efficiency gains but tempered by concerns over clinical decision-making and job displacement.
• 🔬 Training Transformers to solve 95% failure rate of Cancer Trials — Ron Alfa & Daniel Bear, Noetik — Improving cancer drug success requires advanced AI models for precise patient selection based on multimodal biological data.
• 😺 Mayo's AI spotted cancer 3 years before doctors did — AI is demonstrating superior diagnostic and triage capabilities in medicine, even with incomplete information.
• Spotlight on innovation: Google-sponsored Data Science for Health Ideathon across Africa — Google's Ideathon fostered African innovation in healthcare AI using open models like MedGemma and MedSigLIP.
• AI identifies early risk patterns for skin cancer — AI models can predict melanoma risk with high accuracy using routine healthcare data, enabling targeted screening.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Marcus on AI.