Miles Wang, a researcher at OpenAI, is reportedly in discussions to launch an artificial intelligence drug discovery company at a valuation of roughly $2 billion before it has publicly launched a product, a number that captures both the ambition and the froth of the moment when frontier AI talent starts founding companies. The talks would place Wang among a wave of lab insiders betting that the same scaling laws that transformed language can transform biology.
The premise of AI drug discovery is that models trained on molecular structures, genomic data and clinical literature can propose candidate compounds faster and cheaper than traditional screening, compressing a decade-long pipeline into a sequence of computational experiments followed by targeted validation. The field has attracted heavy investment, and recent results from large labs, including protein structure prediction and generative chemistry, have convinced some scientists that the bottleneck is now capital and talent, not proof of concept.
Wang's position inside OpenAI gives him access to exactly the modeling expertise the venture would need, and the reported $2 billion mark reflects investors pricing a founder with front-row exposure to the frontier. It also reflects a broader pattern: the most valuable AI startups of this cycle are increasingly founded by people whose credibility comes from what they built at the major labs.
Skeptics note that AI has yet to deliver a fully approved, commercially transformative drug, and that valuations this large at founding presume a future the science must still earn. If the talks close, Wang's company will carry an unusual burden: proving that a $2 billion promise can survive a decade of clinical trials.
