AI's Rise in Research (Patreon)

For some it is a clear hockey stick moment. But dangers lurk. The rate of change is alarming but are there risks?

AI is rapidly transforming science, with usage in research papers across 20 fields quadrupling in a short window. Note this is going to be a lot higher in 2023 and 2024.

While AI offers powerful tools to speed up discovery, it also brings major risks—especially when used by those without deep technical expertise. Without clear standards and checks, it may flood research with flawed results, giving the illusion of progress while slowing real understanding.

THE GOOD

Artificial intelligence (AI) offers significant advantages in science, such as accelerating drug discovery, enabling efficient knowledge mining, and providing powerful tools for modeling complex systems, while automating repetitive tasks.

THE BAD

However, its misuse and lack of expertise can lead to high error risks, "black-box" predictions without explanations, and issues like data leakage and flawed training methods. Additionally, reproducibility challenges and superficial results can create illusions of progress, potentially undermining scientific theories.

THE UGLY

AI use in medical research —outpaces the scientific process, raising concerns about understanding and responsible implementation.