Nvidia Corp‘s (NASDAQ:NVDA) GPUs have become the backbone of artificial intelligence, but they won’t be enough to determine the winners in AI-driven drug discovery.
In an exclusive email interview with Benzinga, Generate Biomedicines co-founder and CTO Gevorg Grigoryan said advanced computing is a prerequisite for designing AI-generated medicines—but the real competitive edge lies in what companies build on top of that computing power.
Compute Is Essential, but It’s Not the Moat
Grigoryan left little doubt about the importance of access to advanced computing.
“Essential. The scale of model training, molecular generation, and evaluation we perform would not be possible without a large number of GPUs.”
But he quickly drew a distinction between having access to compute and building a sustainable competitive advantage.
“Compute alone is not a differentiator.”
Instead, Grigoryan said the companies most likely to lead AI drug discovery will be those that combine computing with “proprietary biological data, experimental measurement scale, and the ability to close the loop between generation and measurement.”
In other words, GPUs provide the foundation, but proprietary data and continuous learning are what transform that foundation into a durable platform.
Why the Feedback Loop Matters
Generate’s approach is built around using AI to generate new molecular candidates, testing those candidates experimentally and feeding the results back into its models to improve future predictions.
That continuous cycle of generation, measurement and learning allows the company’s AI models to become progressively better over time. According to Grigoryan, simply increasing computing power cannot replicate that advantage without access to unique biological data and large-scale experimentation.
His comments reflect a broader shift in AI investing. As advanced computing becomes more widely available, investors are increasingly looking beyond infrastructure to identify companies with proprietary datasets, differentiated platforms and self-improving AI systems.
Investment Takeaway
Grigoryan’s perspective suggests the competitive landscape in AI drug discovery is entering a new phase. While access to Nvidia GPUs remains essential, it is becoming a baseline requirement rather than a defining advantage.
For investors evaluating AI-driven biotech companies, the more important question may be which platforms can continuously generate, test and learn from proprietary biological data—creating a feedback loop that competitors cannot easily replicate.
Courtesy of Generate Biomedicines
Login to comment