Deep inside of Google's parent company Alphabet, the secret lab, the secret lab, tackles a very bold, sounding promise that sounds like science fiction to “solve all diseases.” The company's Isomorphic Labs is currently preparing to launch its first human clinical trial of cancer drugs fully designed by artificial intelligence.
In a recent interview with Fortune, Colin Murdoch, president of Isomorphic Labs and chief business officer at Google Deepmind, confirmed he is on the crisis of this monumental step. For those who have seen their loved ones fight a devastating illness, the hope this offers is immeasurable. But as we become increasingly wary of the power of AI, it raises a frightening question: can we really trust “black box” algorithms in our lives?
The isomorphic lab was born from Deepmind's famous Alphafold breakthrough, an AI system that surprised scientists by predicting complex 3D shapes of proteins. To understand why this is a big deal, you need to know how drugs are traditionally made. For decades, it has been a slow and brutal process of trial and error. Scientists put one new drug on the market, spending an average of 10-15 years or more, bringing it to the market, with most candidates failing along the way.
Isomorphic Labs use the AI Alphafold 3 to radically accelerate this. AI can predict the complex 3D structure of human body proteins with incredible accuracy, allowing scientists to digitally design new drug molecules that are fully shaped to combat specific diseases.
The company has already signed a multi-billion dollar agreement with drug Novartis and Eli Lily, raising $600 million in new funds to move its own drug candidates to launch oncology in human trials. The promise is medical utopia. “This funding will further turbocharge the development of the next generation of AI drug design engines, helping us move our own program into clinical development, and take an important step towards our day's mission to solve all diseases with the help of AI.”
But when Big Tech begins designing medicine, who owns your treatment? This is where deep fears about the role of AI in our lives focus. The biggest concern is the “black box” issue. I know that AI will give answers, but I don't know how. This raises an important question:
- Does Alphabet own the following cancer drugs to own search results?
- Are these AI-designed treatments affordable or are they trapped behind empty patents that only the wealthy can access?
- Will human testing standards keep up with the enormous speed of machine-generated breakthroughs?
- And who is responsible if AI-designed drugs don't work? A company that owns AI? programmer? AI itself?
When contacted by Gizmodo, a spokesman for Essay Labs said the company “has no more to share.”
AI can revolutionize medicine. But leaving it unchecked can also replicate the worst part of the tech industry: opaque, monopoly and access. The isomorphic lab pushes humanity towards a monumental turning point. If they were successful, they could alleviate more suffering than any other invention in history.
But in order to do so, they must first convince the skeptical public that the promise is worth the unprecedented risk.
