Why Has AI Not Solved Cancer Yet? A Reality Check
In this article
The vast gap between the fantasy that the right algorithm will magically end cancer and the complexity of biological reality.
Over the past few years, cancer has taken some of the scientific world’s brightest minds and thousands of our loved ones from us. It was thought that AI would change all this very quickly. So far, however, that expectation has not been fulfilled. Silicon Valley entered the medical system with an “outsider’s arrogance”, without fully understanding it, and refused to learn from past mistakes.
The three great walls facing AI in the fight against cancer
According to scientists, AI’s inability to solve cancer is due not merely to “inadequate software”, but to formidable barriers inherent in medicine and biology.
Our bodies consist of trillions of interconnected molecules, and every body is different. A drug that works in some people fails in the clinic because we do not understand biology well enough to know “which is which”.
Sometimes a new drug is abandoned because it is “not profitable”, even though it works scientifically. Expiring patents or excessively costly production are economic problems that even perfect AI algorithms cannot solve.
A Critical Statistic: The Gap Between Animals and Humans
With AI, we discover drugs that succeed in animal experiments, but most of these drugs are not found safe in humans. Analyses show that 92% of drug trials that work in animals fail in humans because of toxicity (poisonous effects). AI has not yet fully deciphered the “hidden languages” of human biology.
In what language is AI trained?
As Eli Lilly CEO David Ricks has also emphasised, current AI models are a long way from curing cancer. Why? Because they were trained not to solve questions of biology or chemistry, but to imitate “human language” .
Truly defeating cancer will require both technical and economic systemic changes in how we develop and test drugs. AI must pay much closer attention to details and grasp the logic of the physical world (biochemistry).
What should we expect from AI in the near future?
Rather than expecting AI to make a radically new discovery, in the short term we will see its greatest contribution in the following areas:
- Simplifying everyday tasks and accelerating processes.
- Recruiting patients for clinical trials and managing bureaucratic procedures.
- Providing “advanced tools” that enable people to analyse existing data better.
In short, for now AI plays the role of “an assistant that empowers people”, rather than an “explorer”.
Biological restoration and the challenge of informatics
On our journey towards healing from cancer, we can become captivated by technology’s magic and forget biology’s calm and depth. Gary Marcus’s analysis reminds us that precise guidance is not just about having the fastest processor; it is about training that processor to hear “the cell’s voice”.
The philosophy of biological restoration sees the body and disease not as a pile of “code”, but as a living orchestra in which trillions of variables intertwine. The oncology of the future will be built at that precise crossroads where AI moves from “human language” to the “language of biochemistry” and unites with the physician’s compassion. True healing lies not in technology’s arrogance, but in our humility in understanding nature’s mystery.
Scientific Sources and Further Reading
- 1. Marcus G. "The real test of AI - F Cancer." Substack Analysis. March 16, 2026.
- 2. Javorsky E. "Technosolutionism and the systemic limits of medical progress." Future of Life Institute. 2025.
- 3. Ricks D. "AI training on the language of chemistry and biology." Eli Lilly Corporate Review. 2026.
Was this article helpful?
Your feedback helps us improve our content.
Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.