Over the last half-dozen years of my 25-year career in biotech, I’ve enjoyed more and more moments like this. The vast majority have been thanks to U.S. leadership in drug innovation. But we are at an important crossroads. American patients face high healthcare costs at home, while the rise of China and AI adds new anxieties. As America celebrates its 250^th anniversary, we should reflect on what made this country the world’s engine of drug innovation and how we can sustain it for American patients.

The Innovation Treadmill: America’s Hidden Longevity Machine

America’s leadership in drug innovation was built on a simple bargain: reward new therapies, then force rapid generic substitution once patents expire. Consider Pfizer’s Lipitor, a leading statin that reduces cardiovascular mortality by roughly 25%. Within months of losing exclusivity, generic substitution reduced Lipitor’s price by 90%. 

Lipitor’s story applies to nearly every drug. Companies have, on average, 14 years to recover over a billion in development costs before products lose exclusivity. That keeps them racing to generate the next breakthrough. We call this dynamic the “Innovation Treadmill”.

The Innovation Treadmill needs to provide sufficient incentives for innovators to pursue novel therapies, which can make the ~10% of branded drugs expensive. When insurance provides coverage, the average American’s share of $120 per month provides access to all medicines past and present. As a result of the Innovation Treadmill, the U.S. has dominated global drug innovation. American patients get early access to the most novel medicines. In contrast, European countries are losing ground, with access to 70% or less of novel drugs, often with delays of a year or more. 

Europe’s decline comes at a terrible time for its people. We are entering a golden age of innovation, enabled by four mega-trends: cheap genetic information, a drug modality boom, artificial intelligence, and China’s rise. 

The Genetic and Modality Boom 

The Human Genome Project cost nearly $3 billion and took over a decade. Today, genome sequencing costs around $200, making it possible to identify and target the genetic roots of disease. Cell and gene therapies are genomic medicine’s most powerful expression, holding promise for the roughly 7,000 known rare diseases. Since the first FDA-approved gene therapy in 2017, the field has produced over 30 novel therapies. 

Cell and gene therapies are only two of several new modalities that have more than tripled the ways diseases can be treated. For example, over the next decade, substantially all chemotherapy is likely to be made obsolete by antibody drug conjugates (ADCs), which deliver therapy more selectively to tumors. Enhertu and Padcev are examples of ADCs that have extended survival by 10 and 15 months in breast and bladder cancer, respectively. 

And perhaps nothing is better known today than GLP-1s, which demonstrate the transformative potential of peptides. Like biologics, peptides offer high target precision with the added advantage of oral delivery. Icotyde, approved in March, makes the most successful injectable mechanism in immunology (used by Skyrizi and Tremfya) available in pill form for the first time. 

China and AI are Innovation Turbochargers

America is no longer fighting the battle against disease alone. Over a decade, China built its own Innovation Treadmill, by pushing down generic prices while expanding coverage for novel medicines. China’s ecosystem also has some unique advantages: twice as many scientists, larger patient pools for clinical research, and world-leading manufacturing. China now produces roughly a third of new drug candidates.  

The impact of AI is likely to be even greater. AI can boost productivity across three critical areas: target discovery, drug design, and clinical trial efficiency. The first two play to AI’s natural advantages in handling large, unstructured datasets. AI can mine vast libraries of genetic data, molecules, and predict structures and drug interactions. On the clinical side, AI’s benefits are more mundane but no less impactful. Of the roughly seven years a drug spends in clinical trials, two are spent on paperwork, and more than half of the remainder by finding patients for studies. AI can accelerate both, and America is well positioned to lead.

Protecting and Accelerating the Innovation Treadmill 

These mega-trends are wonderful reasons to be hopeful about our future health. Humanity’s battle against disease has only just begun, and we should welcome every nation and AI to the effort.

But progress is not inevitable. Inflation in hospital and services costs has driven total healthcare spending to over $1200 per month. It continues to rise rapidly, causing the majority of Americans to cry out for relief. So far unable to contain services inflation, insurers and policymakers have continued to focus on modifying drug spending. Policies that undermine innovation, such as the Pill Penalty within the IRA, and a brief flirtation with anti-innovation leadership at the FDA, showed how fragile and sensitive to change the grand bargain can be. 

As we mark 250 years of American ingenuity, this golden age of innovation needs the U.S. to continue to lead. America needs to modernize with the same ambition that built the Innovation Treadmill in the first place. We should streamline the path from lab to patient by reducing unnecessary animal testing. We should expand master protocols beyond oncology to eliminate lost time between studies. And we should normalize regulatory flexibility and risk-based frameworks for serious unmet needs and immature new modalities. Finally, we should leverage AI and electronic health records to speed up patient recruitment and administrative work. 

Through these reforms, we ensure our Innovation Treadmill can keep pace with accelerating science and continue to deliver transformative medicines to American patients first.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.