In an era defined by rapid technological advancements, artificial intelligence (AI) is revolutionizing numerous industries, and its impact on drug discovery is proving to be particularly transformative. The emergence of companies like Lila Sciences and Recursion Pharmaceuticals highlights a growing confidence among investors and researchers that AI can unlock unprecedented scientific insights, accelerating the development of new drugs and reshaping the landscape of scientific exploration.
Lila Sciences, with its ambitious vision of creating “scientific superintelligence,” and Recursion Pharmaceuticals, which leverages an AI-powered platform to map human biology, are at the forefront of this paradigm shift. Both companies have secured substantial funding, hundreds of millions of dollars, and are utilizing the latest advancements in AI scaling laws to position themselves for major breakthroughs in medicine, materials science, and other fields. The advancements are based on distinct scaling laws, similar to Moore’s Law.
Rise of Scientific Superintelligence
Lila Sciences combines generative AI with a network of autonomous laboratories, where AI systems design, test, and refine scientific hypotheses in real-time. The company aims to build a self-reinforcing loop in which AI continuously generates and tests new ideas, thereby accelerating the scientific method and achieving discoveries that would be unattainable by human researchers alone.
According to Geoffrey von Maltzahn, co-founder and CEO of Lila, “Our hypothesis is that by scaling experimentation, we can unlock emergent abilities and enable discoveries that remain hidden at smaller scales.”
Historically, scientific progress has followed a methodical, human-limited path: hypothesize, experiment, analyze, and repeat. This approach has led to remarkable findings. However, the vastness of potential chemical, biological, and physical interactions limits even the most brilliant minds from exploring more than a fraction of possibilities. Lila Sciences, established in 2023 within Flagship Pioneering’s innovation labs, seeks to overcome these limitations by developing “scientific superintelligence” – advanced AI systems that can not only process existing scientific data but also autonomously generate hypotheses, design experiments, and extract meaningful insights at scales impossible for human scientists.
Early results from Lila have been promising, with applications in materials science already evident. The company has developed catalysts for green hydrogen production, as well as materials for carbon capture – critical technologies for addressing climate change. Similarly, Recursion has developed a processing pipeline and neural network platform that has identified potential treatments across various disease categories, producing a robust pipeline of drug candidates.
AI-Driven Map of Human Biology
While Lila Sciences focuses on scientific superintelligence, Recursion Pharmaceuticals, founded in 2013, has been building an AI-enabled map of human biology.
Recursion combines experimental biology, bioinformatics, and machine learning to identify potential treatments for diseases at an unprecedented scale and speed. The company’s platform integrates automated biology, chemistry, and cloud-based computing to test thousands of compounds simultaneously. This strategy aims to circumvent “Eroom’s Law”, the counterintuitive observation that the cost and time required to bring new drugs to market have increased despite technological advancement. Recursion seeks to reverse this trend by using AI to automate and accelerate drug discovery’s early stages.
Recursion’s AI models analyze cellular-level data to identify patterns and predict compound interactions with biological systems. By constructing a comprehensive map of human cellular biology, Recursion seeks to uncover novel drug targets and therapeutic strategies more efficiently and cost-effectively than traditional methods. Chris Gibson, CEO of Recursion, explains, “Recursion is not just trying to find the next drug; we’re trying to redefine how drugs are discovered altogether. The combination of AI and large-scale biological data has the potential to unlock entirely new categories of medicine.”
Three Scaling Laws Driving AI Advancement
What makes companies like Lila and Recursion possible today – rather than a decade ago – is our deepening understanding for how AI systems scale and improve. Three critical scaling laws now guide development in the field:
Pre-Training Scaling
The first scaling law shows increasing intelligence and accuracy in larger models trained on more data with greater computational resources. This principle has driven the development of billion- and trillion-parameter transformer models that form the backbone of modern AI systems. For scientific applications, this means AI systems can now ingest and process the entirety of scientific literature – millions of papers, experimental results, and theoretical models – creating a knowledge base far beyond what any individual scientist could master.
Post-Training Scaling
Once foundation models are pre-trained, they can be honed for specific domains through fine-tuning, pruning, quantization, and distillation. Andrew Beam, Ph.D., CTO of Lila Sciences, notes, “The post-training ecosystem of derivative models could require around 30 times more compute than training the original foundation model. This significant computational investment allows us to create models specifically optimized for different scientific domains.”
For drug discovery companies, this translates to the creation of specialized models that can understand protein folding, molecular interactions, cellular biology, and chemical synthesis, each requiring domain-specific training based on general scientific knowledge.
Test-Time Scaling (Long Thinking)
Perhaps most revolutionary for scientific applications is test-time scaling which allows AI systems to reason through complex problems during inference instead of providing immediate answers. Kenneth Stanley, Ph.D., Senior Vice President at Lila Sciences, explains:” On challenging scientific questions, this reasoning process might take minutes or even hours, requiring over 100 times the compute of traditional AI inference. The result is a much more thorough exploration of potential solutions, similar to how human scientists would approach a complex problem.”
This capability enables AI to break down complex scientific questions, explore multiple potential solutions, and show its reasoning – a critical feature for scientific applications where transparency in the discovery process is essential.
The Talent Behind the Revolution
Success in this space requires an interdisciplinary team with expertise in AI, biology, chemistry, and robotics. Lila Sciences has assembled an impressive team including renowned geneticist George Church, Ph.D.; AI expert Andrew Beam, Ph.D.; and AI research pioneer Kenneth Stanley, Ph.D., known for his work on neuroevolution and open-ended algorithms. Recursion similarly boasts an interdisciplinary team combining expertise in experimental biology, machine learning, and drug development, allowing them to bridge the gap between computational predictions and laboratory validation.
The Future of AI in Science
As AI models continue to grow in complexity and capability, the competitive landscape in drug discovery and scientific research is likely to shift. Companies that can harness AI scaling laws and build autonomous experimentation platforms will have a distinct advantage in discovering novel treatments, materials, and energy solutions.
Lila Sciences and Recursion Pharmaceuticals exemplify two complementary approaches to this formidable challenge. Lila’s focus on scientific superintelligence positions it to drive breakthroughs across multiple domains, while Recursion’s deep understanding of biology and drug discovery gives it a strategic edge in developing new medicines. The race to build scientific superintelligence is only beginning. However, if the early success of Lila and Recursion is any indication, AI-driven platforms could soon unlock discoveries that redefine human health, energy production, and scientific understanding itself.
