Big tech says AI can find a cure for cancer. So where is it?

Artificial intelligence is being hailed as a potential saviour for cancer treatment, with big tech executives boldly proclaiming that AI will find a cure within five to ten years or even within our lifetimes. However, the reality of AI's impact on cancer treatment is more nuanced. While AI has made significant progress in drug development, treatment plans, and clinical applications such as imaging for early signs of malignancy and assisting in brain tumour removal, there are still many challenges and limitations.

Cancer is a complex and diverse group of diseases, with over 100 types of tumours in the brain alone, and outcomes depend on various factors such as the biology of the disease and available treatments. A single cure for cancer is highly unlikely. Although AI can improve several aspects of the cancer treatment process, including identifying risk factors, detecting disease, characterizing tumours, selecting treatments, and monitoring patient progress, it is not sufficient to find a cure for all cancers in the next decade.

Medical imaging, in particular, stands to benefit from AI and machine-learning models. Training models on large amounts of patient data enables them to detect changes or patterns that may be imperceptible to the human eye. This has led to advancements in early detection of pancreatic cancer, where machine-learning tools can identify microscopic changes in pancreases that appear normal during clinical diagnosis.

However, further studies are needed to determine the robustness and reproducibility of these methods across different institutions.

While AI can contribute to more personalised and equitable cancer care, it is essential to acknowledge the tension between the grandiose claims of big tech and the deliberate, considered progress of science. As Sherene Loi, a medical oncologist, points out, until AI can think and conduct experiments independently at scale, the prospect of finding a cure for all cancer or medical illness within the next ten years remains uncertain.