AI revolutionizes medicine: creates drugs in record time

Just a decade ago, the thought of a computer designing drugs seemed like something out of a futuristic movie. Today, that fiction is a tangible reality. Artificial intelligence (AI) has begun to radically transform drug development , shortening timelines, reducing costs, and offering unprecedented precision.

Traditionally, developing a drug can take more than 10 years and cost up to $2 billion. It's a long road, fraught with trial, error, and clinical hurdles. But the landscape has changed.

Companies like Insilico Medicine and Benevolent AI are already using algorithms to analyze vast chemical and genomic databases , finding promising compounds in just weeks. One of the most impactful cases occurred in 2024, when Generate Biomedicines designed an antiviral protein in just 21 days to combat new SARS-CoV-2 variants. The scientific world was astounded.

The turning point came in 2020. That year, AlphaFold , an AI developed by DeepMind (a Google subsidiary), managed to predict the structure of almost all known human proteins with extremely high accuracy. This milestone is considered one of the most significant advances in modern biology, as proteins are key to understanding and treating diseases.

Thanks to AlphaFold, scientists can now visualize how molecules interact in the body , enabling the development of more effective and targeted treatments.

In addition to accelerating drug development, AI is helping to develop treatments tailored to each patient . In oncology, for example, algorithms can analyze a tumor's genetic profile and suggest specific drug combinations. This reduces uncertainty and increases the likelihood of therapeutic success.

“We're moving from a trial-and-error model to a predictive one,” explains Dr. María López , a bioinformatician at the Institute of Innovative Medicine. “AI tells us which molecules have the greatest potential for each type of patient.”

Not everything is optimistic. The rapid advancement of this technology also raises questions. What happens if the data used to train the algorithms is biased ? This could affect specific population groups and lead to less effective treatments.

Furthermore, who is responsible if an AI-developed drug produces serious side effects? Experts agree that human intelligence remains indispensable to oversee, validate, and contextualize each result.

Dr. Carlos Ruiz , an Argentine pharmacologist, sums up the feelings of many researchers:

“AI won't replace us, but it's like a magnifying glass on steroids. We can see more, understand better, and act faster.” And that's precisely what 21st-century medicine needs: speed, precision, and personalization .