The Living Algorithm: How Biological Feedback Loops Will Unblock the AI Drug Discovery Bottleneck
The last few years have seen undeniable, headline-grabbing breakthroughs in computational biology. With generative AI dreaming up novel proteins and deep learning solving decades-old structural mapping challenges, the industry has sold a compelling narrative: that we are on the cusp of fully in silico drug design. But these advances have birthed a pervasive misconception—the idea that the primary bottleneck in therapeutics is simply a lack of predictive algorithmic power. It isn’t. The foundational problem in drug discovery, whether traditional or AI-driven, is that the process remains stubbornly sequential. The industry currently treats AI as a standalone oracle, applying massive models in isolated, disconnected steps that leave ultimate clinical viability entirely up to chance. To move beyond the hype, we have to stop treating biologics as a one-shot prediction game. The true frontier isn’t just scaling up compute; it is shifting the entire paradigm from sequential guessing to continuous learning.
The evening will feature dinner, discussion, and networking with entrepreneurs at the intersection of AI and health care.
Ido Calman
CEO and Co-founder
Onava
