Researchers and biotech companies are increasingly exploring artificial intelligence methods to accelerate drug development. Simon Kohl, former AlphaFold2 co-developer and current CEO of Latent Labs, has launched LatentX—a platform aiming to design custom biological molecules with high effectiveness. LatentX claims to reach laboratory hit rates of up to 100 percent for macrocycles, a substantial jump over traditional drug discovery’s minimal success rates. As the boundaries between biology and computation blur, scientists are optimistic about reducing time and costs in creating new treatments for diseases that have been challenging to target. Despite these advances, questions remain about the implications of rapidly generating novel biological entities, as both opportunities and responsibilities expand.
Kohl’s work builds on innovations that once made headlines with DeepMind’s AlphaFold2, which accurately predicted protein structures but did not design new molecular forms. Recent years saw AlphaFold2 being leveraged in academic and pharmaceutical circles for understanding existing biological mechanisms, but actual creation of new biomolecules typically occurred through much slower, experimental approaches. Latent Labs differentiates itself by generating both sequence and structure of proteins from scratch, not just predicting, but actively designing potential drugs—stepping significantly beyond what prior technologies provided. The move from computational prediction to creation marks a distinctive shift in how researchers may soon approach disease treatment.
How Does LatentX Set Itself Apart from Previous AI Technologies?
Unlike AlphaFold2, which focused on interpreting natural proteins, LatentX uses generative AI to author entirely new molecules and optimize them in real time, combining theoretical biochemistry with high-speed computation. This method yields laboratory results where a fraction of the candidates tested translate into real-world, functional compounds. Simon Kohl notes,
“We’re authoring biology, not just predicting it.”
The workflow shortens the timeline from months of laboratory effort to seconds of computational analysis, giving scientists a faster route to testable medicines.
What Areas is Latent Labs Prioritizing?
Latent Labs initially targets therapeutic areas where existing drug discovery methods often yield poor results, such as oncology, rare genetic disorders, and autoimmune conditions. Macrocycles, an especially challenging drug format, are a primary focus due to their merging of biologic precision and drug-like properties. Investors from both the technology and bioscience sectors, including Jeff Dean and Aidan Gomez, have backed the company’s platform, believing it could address previously intractable medical problems. The potential commercialization of such custom-designed therapeutics is seen as a strategic advantage in the rapidly expanding biologics market.
How Does Latent Labs Address Safety Concerns Regarding AI-Generated Molecules?
Aware of the risks linked to creating new biological systems, Latent Labs collaborates with regulatory bodies and integrates biosafety protocols within its platform. The company restricts user access according to international standards and validates all computational outputs through physical laboratory experiments in San Francisco. According to Kohl,
“We take dual-use implications seriously—actively participating in biosafety discussions with regulators and restricting access per international sanctions lists.”
By merging computation with hands-on wet lab validation, Latent Labs aims to better understand and control the ramifications of its technological advances.
Latent Labs’ approach reflects a wider shift in drug discovery, as companies move from prediction to creation. Scientists and engineers are now expected to consider not just efficacy but also the broader societal impact of their work. The discussion on how to balance innovation, accessibility, and safety continues, especially as the tools for creating novel biological systems become more available. For researchers and biotech companies, AI-powered platforms like LatentX are likely to become increasingly central to early-stage therapeutic development. Readers should note that while these tools could accelerate research and introduce new drugs to market faster, true safety and effectiveness must still be proven through traditional laboratory and clinical processes. As AI-generated biology gains momentum, transparency, governance, and collaboration with regulators will be critical in managing both the promise and risk of digital-era biotechnology.
