I think we’ve all felt that unique kind of helplessness, right? You hear about a friend, a family member, or maybe even yourself getting a scary diagnosis, cancer, an autoimmune disease, something terrifying.
And your first thought is, “Okay, what’s the cure? What’s the treatment?” Then you run headfirst into the brutal reality of medicine: developing a new drug takes, on average, over a decade and costs billions of dollars. And even then, most of them fail.
It’s a slow, grinding process of trial and error that feels hopelessly outdated in our age of instant everything. It’s a frustration I’ve felt deeply. But what if we could change the entire game? What if we could stop guessing and start designing?
Well, that science-fiction future just took a giant leap into reality. Google DeepMind’s sister company, Isomorphic Labs, just confirmed they are getting ready to start HUMAN TRIALS for drugs that were designed from the ground up by artificial intelligence. This is not a research paper or a thought experiment. This is happening.
Colin Murdoch, the head of Isomorphic Labs, put it plainly: “There are people sitting in our office in King’s Cross, London, working, and collaborating with AI to design drugs for cancer. That’s happening right now.” The next step? Putting those AI-designed molecules into people to see if they work. It’s mind-blowing.
🤯 How Did We Even Get Here? The AlphaFold Story
To understand why this is such a monumental deal, we have to talk about the AI that started it all: AlphaFold.
Your body is made up of trillions of cells, and inside those cells are billions of tiny biological machines called proteins. They do literally everything: carry oxygen, fight invaders, send signals, build tissues. The function of a protein is determined by its intricate, unique 3D shape.
Think of it like a key: its shape determines which lock it can open.
For decades, figuring out the exact 3D shape of a single protein was a nightmare. It could take a scientist their entire PhD, years of painstaking work in a lab, to map out just one. This was the single biggest bottleneck in biology. If you don’t know the shape of the protein that’s causing a disease (the “lock”), how can you possibly design a drug (the “key”) to fit it?
Then, in 2020, DeepMind unleashed AlphaFold. This AI was trained on all the known protein structures and learned the fundamental rules of how they fold. It can now predict the 3D structure of almost any protein with jaw-dropping accuracy, in a matter of minutes. It solved one of the grand challenges of biology. This wasn’t just an improvement; it was a revolution. The creators of AlphaFold even shared a Nobel Prize for it in 2024, that’s how big this is.
Isomorphic Labs was born directly from this breakthrough. Demis Hassabis, the genius behind DeepMind, realized that if AI could predict the shape of these biological locks, then the next logical step was to use AI to design the perfect keys. And that’s exactly what they’re doing.
⚙️ The Old Way vs. The AI Way: A Total Game-Changer
The difference between traditional drug discovery and what Isomorphic is doing is like the difference between searching for a needle in a haystack and 3D-printing a custom needle on demand.
📌 The Old Way (The Brutal Slog)
- The Target: Scientists identify a protein that’s misbehaving and causing a disease.
- The Search: They then take tens of thousands of existing chemical compounds off a shelf and, one by one, test them against the protein to see if anything sticks. It’s a brute-force approach.
- The Analogy: Imagine you have a new, complex lock but you’ve lost the key. Your only strategy is to go to a warehouse with millions of random keys and try every single one until you find one that happens to fit, or sort of fits.
- The Result: It’s incredibly inefficient. The process takes years, costs a fortune, and has a failure rate of over 90%. So many promising ideas die on the vine simply because finding the right starting molecule is too hard.
🚀 The New AI Way (Intelligent Design)
- The Blueprint: Thanks to AlphaFold, Isomorphic starts with a perfect, high-resolution 3D model of the disease-causing protein (the lock).
- The Design: Instead of searching for a key, their AI designs a brand new key from scratch. It analyzes the precise shape, pockets, and grooves of the protein target and computationally generates a molecule that is tailor-made to fit it perfectly.
- The Speed & Precision: This virtual design process shrinks years of lab work into weeks or even days. Because the drug is designed for its target, it has a much higher probability of being effective and a lower chance of causing unwanted side effects (i.e., fitting into the wrong locks).
- The Vision: As Murdoch said, the ultimate goal is to one day say, “Here’s a disease,” and have an AI model spit out the design for a drug to treat it. We’re not there yet, but this is the first real step on that path.
✨ What This Actually Means For You & Me
Okay, this is awesome tech, but what’s the real-world impact? This is where it gets incredibly exciting.
Isomorphic Labs is initially focusing on some of the toughest areas in medicine, including oncology (cancer) and immunology (autoimmune diseases). Think about diseases like aggressive cancers, rheumatoid arthritis, or Crohn’s disease, where current treatments can be harsh and often aren’t effective for everyone.
AI-designed drugs could lead to:
- More Effective Treatments: By creating drugs that are a perfect fit for their target, we can hit the disease harder and more accurately.
- Fewer Side Effects: Precision targeting means the drug is less likely to interfere with healthy proteins in the body, which is the cause of many debilitating side effects.
- Personalized Medicine: In the future, it’s conceivable that we could sequence a tumor from a specific patient, model its unique proteins, and have an AI design a drug just for them. This is the holy grail of medicine.
- Tackling “Undruggable” Diseases: Some diseases are caused by proteins that were considered “undruggable” because their shapes made them impossible targets for traditional methods. AI can analyze these complex structures and find new ways to attack them.
This isn’t just about making the existing process a bit better. It’s about fundamentally rewiring how we fight disease. The partnerships Isomorphic has already inked with pharmaceutical giants like Novartis and Eli Lilly, plus the $600 million in funding they’ve secured, prove that the entire industry knows this is the future.
✍️ The Road Ahead Is Long, But The Journey Has Begun
Let’s keep our feet on the ground for a second. Getting to human trials is a massive milestone, but it’s the start, not the finish line. Clinical trials are a long, multi-phase process designed to ensure a new drug is safe and effective.
- Phase I: Tests for safety in a small group of healthy volunteers.
- Phase II: Tests for effectiveness and side effects in a small group of patients.
- Phase III: Large-scale testing on hundreds or thousands of patients to confirm effectiveness and monitor long-term safety.
This process still takes years. But the critical difference here is the starting point. Instead of entering this expensive and lengthy gauntlet with a molecule that has a 90% chance of failure, they’re starting with one that was intelligently designed from first principles. The odds are already tilted in our favor.
What we are witnessing is a pivotal moment where the two most complex systems we know: human biology and artificial intelligence, are finally coming together to solve our biggest challenges. It’s the beginning of a new era, and I, for one, can’t wait to see what diseases we cure next.
The technology at the heart of Isomorphic Labs is AlphaFold, an AI system from Google DeepMind that can accurately predict the 3D structures of proteins. This breakthrough allows scientists to understand how diseases function at a molecular level and to digitally design drugs that can precisely interact with specific protein targets.
To accelerate its mission, Isomorphic Labs has forged major partnerships with pharmaceutical giants Novartis and Eli Lilly. These collaborations, alongside a $600 million funding round in April 2025, are helping to advance the company’s internal drug pipeline, with an initial focus on oncology and immunology.
While Isomorphic Labs is a pioneer, it is part of a broader industry trend. Thousands of drugs developed or repurposed using AI are currently in early development stages worldwide. The success of Isomorphic’s upcoming clinical trials could serve as a major validation for the field, potentially shortening drug development timelines from years to months and lowering costs.