While ChatGPT and other large language models dominated AI headlines throughout 2024, a quieter revolution has been transforming pharmaceutical laboratories worldwide. Industry leaders are already looking beyond the chatbot era: only 31% of respondents see LLMs as the most significant AI development for the coming year. In contrast, 65% are betting on AI agents – autonomous systems that analyze data, make decisions, and take action across complex workflows – as the breakthrough that will define 2025.
“65% of industry leaders think AI agents will be the defining breakthrough of 2025”
This shift from conversation to creation is already reshaping drug discovery. At SuperAI Singapore, the world’s largest AI event, Alex Aliper (Co-founder and President of Insilico Medicine) offered a window into how this transformation is unfolding. Speaking with Bloomberg on the sidelines of the event, Aliper revealed how AI in pharma has evolved far beyond simple pattern recognition.
Today's systems are designing molecules, predicting clinical outcomes, and navigating the complex global infrastructure required to bring new medicines to market. While consumers interact with AI predominantly through chatbots, pharmaceutical researchers are applying AI technologies to design medicines that don't yet exist. As Aliper put it, the mission is nothing less than "extending healthy productive longevity for everyone on the planet”.
From conversation to creation
Where early machine learning applications in drug discovery focused on mining existing data for patterns and predicting which known compounds might work for new diseases, today's systems actively generate novel molecular structures that have never existed before. This evolution from passive analysis to active creation is what separates the current wave of AI in healthcare from previous computational approaches. Insilico Medicine demonstrated this shift in a Nature Biotechnology paper earlier this year, reporting the first experimentally validated molecules generated using quantum computing. Working with IBM's 16-qubit quantum computer and the University of Toronto, the team created potential drug compounds, tested them in oncology programs, and proved they worked. "This is the first study to date that demonstrates that quantum can be used to generate experimentally validated structures," Aliper explained. This success is rooted in a crucial insight, one which 55% of SuperAI’s PULSE survey respondents share: AI is "a tool, not the point." Rather than pursuing artificial intelligence for its own sake, leading-edge pharma companies like Insilico are integrating these technologies end-to-end across the entire drug development pipeline.
Looking ahead, Aliper's commitment to invest "more into R&D, both AI and wet lab," captures the industry's direction: not full automation but deeper integration of human expertise and machine learning.
AI is quietly working in pharmaceutical laboratories to extend human life. This represents the practical application of machine learning to one of humanity's oldest challenges: fighting disease. This revolution will be published in peer-reviewed journals, validated in clinical trials, and ultimately measured in lives saved.
The quiet nature of this transformation reflects the industry's character. Drug discovery has always been a field where results matter more than rhetoric, and where a single successful therapy justifies years of patient work.
“AI's most practical phase is just beginning”
The PULSE survey concludes that "AI's most practical phase is just beginning," and pharma R&D is showing us exactly what that looks like. Not hype cycles or viral moments, but systematic integration of AI agents into every stage of development, supported by massive infrastructure investments and guided by the principle that healthcare is a universal human right.
For those watching AI's evolution, Insilico's journey offers crucial lessons. Success requires more than algorithms: it demands infrastructure, global collaboration, and a clear focus on tangible outcomes. It requires viewing AI as Aliper and the majority of industry leaders do: as a tool for achieving human goals, not an end in itself.
As this quiet revolution continues, the breakthrough will be measured in the longevity and quality of human life – the ultimate test of any technology's worth.
About the author
