1
people have died from curable diseases
since this page started loading...
💀

The Untapped Therapeutic Frontier

Keywords

war-on-disease, 1-percent-treaty, medical-research, public-health, peace-dividend, decentralized-trials, dfda, dih, victory-bonds, health-economics, cost-benefit-analysis, clinical-trials, drug-development, regulatory-reform, military-spending, peace-economics, decentralized-governance, wishocracy, blockchain-governance, impact-investing

You’ve explored less than 1 percent of medicine. The other 99 percent is where the cures are.

You’ve explored less than 1 percent of medicine. The other 99 percent is where the cures are.

You think you’ve explored medicine. In reality, you’re standing in the parking lot of Disney World bragging about how fun the pavement is. You’ve taken pictures of the asphalt. You’ve written dissertations about the parking stripes. You’ve never gone inside.

We’ve Studied Less than 1 percent of Potential Safe Treatments

We’ve Studied Less than 1 percent of Potential Safe Treatments

After a century of pharmacology, most beneficial effects of molecules on diseases remain unexplored. Not because biology is constrained. Because clinical trials are slow, expensive, and run by people who think “urgency” is a type of perfume.

Here’s some math. Don’t worry, it’s painful but brief.

Your Tiny Sandbox

Here is everything humanity considers “medicine” after 5,000 years of civilization:

A visualization showing the scale difference between the 20,000 commercial drug products and the much smaller pool of ~9,500 unique safe-ish compounds that comprise humanity’s entire medicinal knowledge.

A visualization showing the scale difference between the 20,000 commercial drug products and the much smaller pool of ~9,500 unique safe-ish compounds that comprise humanity’s entire medicinal knowledge.

Total unique compounds humanity knows are safe-ish: 9.50k compounds (95% CI: 7.00k compounds-12.0k compounds).

That’s it. That’s your entire medicine cabinet after five millennia.

The Target List (Ways Your Body Breaks)

  • Codes: The ICD-10 has 14.0k codes codes for ways your meat can malfunction.
  • Real Targets: Consolidation gives 1.00k diseases (95% CI: 800 diseases-1.20k diseases) trial-relevant diseases worth fixing.

A funnel diagram illustrating the consolidation of 14,000 ICD-10 medical codes down to approximately 1,000 trial-relevant disease targets.

A funnel diagram illustrating the consolidation of 14,000 ICD-10 medical codes down to approximately 1,000 trial-relevant disease targets.

The Math You’re Ignoring

Take your 9.50k compounds (95% CI: 7.00k compounds-12.0k compounds) safe compounds. Multiply by 1.00k diseases (95% CI: 800 diseases-1.20k diseases) diseases. This gives you the “Combinatorial Space,” which is fancy talk for “stuff you could test tomorrow if you weren’t busy filling out forms.”

A visualization illustrating the scale of the combinatorial space, showing how the intersection of thousands of compounds and diseases results in millions of potential testing combinations.

A visualization illustrating the scale of the combinatorial space, showing how the intersection of thousands of compounds and diseases results in millions of potential testing combinations.

9.50k compounds (95% CI: 7.00k compounds-12.0k compounds) compounds × 1.00k diseases (95% CI: 800 diseases-1.20k diseases) diseases = 9.50M combinations plausible drug-condition combinations.

What You’ve Actually Tested

  • Approved uses: 1.75k pairings (95% CI: 1.50k pairings-2.00k pairings) unique approved drug-disease pairings.
  • Repurposed uses: 30% of drugs gain at least one new indication after initial approval.
  • Failed trials: Let’s be generous and say you tested 10 times as many things that failed.

A visual breakdown of drug testing outcomes showing the disproportionate scale of failed trials compared to approved and repurposed drug-disease pairings.

A visual breakdown of drug testing outcomes showing the disproportionate scale of failed trials compared to approved and repurposed drug-disease pairings.

Total tested: 32.5k relationships (95% CI: 15.0k relationships-50.0k relationships) relationships.

The Scorecard

32.5k relationships (95% CI: 15.0k relationships-50.0k relationships) tested / 9.50M combinations possibilities = 0.342% (95% CI: 0.21%-0.514%)

\[ \begin{gathered} Ratio_{explore} = \frac{N_{tested}}{N_{combos}} = \frac{32{,}500}{9.5M} = 0.342\% \\[0.5em] \text{where } N_{combos} \\ = N_{safe} \times N_{diseases,trial} \\ = 9{,}500 \times 1{,}000 \\ = 9.5M \end{gathered} \]

Humanity has explored less than 1% of the theoretically testable drug-disease space.

You are missing 99.7% (95% CI: 99.5%-99.8%) of the picture.

A visualization of the drug-disease exploration space, showing a tiny sliver representing the 0.3 percent currently known versus the massive 99.7 percent area that remains unexplored.

A visualization of the drug-disease exploration space, showing a tiny sliver representing the 0.3 percent currently known versus the massive 99.7 percent area that remains unexplored.

It’s like reading the first page of War and Peace and writing a book report called “It’s About a Party.” Technically accurate. Cosmically insufficient.

Is The Untested Stuff Useful?

You might argue, “Maybe the other 99% is useless!”

A visualization showing the 12 percent vs 88 percent split of targeted human biology alongside a network diagram illustrating how drugs interact with multiple overlapping disease pathways.

A visualization showing the 12 percent vs 88 percent split of targeted human biology alongside a network diagram illustrating how drugs interact with multiple overlapping disease pathways.

Biology disagrees.

  • Undrugged Targets: Mapping 350,000+ clinical trials showed that only 12% of the human interactome has ever been targeted by drugs. You’re ignoring 88% of your own biology. It’s like having a 100-room mansion and only ever entering the bathroom.
  • Polypharmacology: Drugs are messy. They hit multiple targets. FDA-approved drugs often bind to things you didn’t intend. This means “side effects” are often “cures for something else” that you haven’t noticed yet. Every headache pill might cure Alzheimer’s. You won’t know until you check.
  • Network Overlap: Diseases cluster on shared biological networks138. A drug for arthritis might cure Alzheimer’s. You won’t know until you check. But you’re not checking.

The Universe is Big. You Are Small.

So far we only counted molecules humanity already has. If you look at what’s chemically possible, it gets embarrassing.

A scale comparison showing the massive disparity between the billions of known molecules and the vast, unexplored chemical space, visualized as a single grain of sand against the backdrop of the entire planet Earth.

A scale comparison showing the massive disparity between the billions of known molecules and the vast, unexplored chemical space, visualized as a single grain of sand against the backdrop of the entire planet Earth.
  • Chemical Space: Estimated 10^23 to 10^60139 drug-like molecules.
  • Virtual Libraries: ZINC-22 contains tens of billions140 of purchasable virtual molecules.
  • Synthesized: You have made <10^(-40) of possible molecules.

Tested / Possible = ~10^-20 to 10^-50.

Effectively zero. You haven’t even started.

This is like exploring Earth by looking at one grain of sand on one beach in New Jersey, then declaring you’ve “seen the world.” You haven’t. You’ve seen a grain of sand. In New Jersey.

Why You Are Still Sick

It’s not because science is hard. It’s because:

  1. Trials Cost Too Much: A typical Phase II-III RCT costs $30-$100M (Median cost per patient $41K (95% CI: $20K-$120K)). At that price, you only test “sure things.” You test the equivalent of betting your house that the sun will rise. Safe bets. Boring bets. Bets that don’t cure cancer.

A conceptual map of the systemic barriers preventing medical breakthroughs, highlighting how high costs, risk aversion, and patent requirements create a bottleneck between scientific potential and patient care.

A conceptual map of the systemic barriers preventing medical breakthroughs, highlighting how high costs, risk aversion, and patent requirements create a bottleneck between scientific potential and patient care.
  1. Herd Mentality: Trials overwhelmingly test the same few biological targets due to preferential attachment dynamics141. Everyone studies what everyone else is studying because that’s how you get grants. It’s like if every explorer only went to France because France already had good reviews on TripAdvisor.

  2. No Profit: Most molecules have no patent. No patent = no $100M trial = no approval. Your potential cure for cancer might be aspirin plus vitamin D, but nobody will ever test it because you can’t patent aspirin plus vitamin D. This is a system designed by people who hate you.

  3. Neglect: If you have a rare disease with no market, you don’t get a drug. Sorry. Your cells malfunction in an unprofitable way. The invisible hand of the market has determined your death is not worth preventing.

How Long to Explore Everything?

At the current pace of 3.30k trials/year (95% CI: 2.64k trials/year-3.96k trials/year) trials per year, how long would it take to systematically test therapeutic possibilities?

Tier 1: Single Compounds (Most Conservative)

Testing every known safe compound against every major disease:

9.50M combinations combinations ÷ 3.30k trials/year (95% CI: 2.64k trials/year-3.96k trials/year) trials/year = 2.88k years (95% CI: 2.45k years-3.43k years)

That’s almost 3,000 years to explore compounds we already know are safe. By that math, we won’t finish until the year 5000.

A visualization of the massive backlog of 9.5 million compound-disease combinations funneling through a narrow annual trial capacity, illustrating why the process will take until the year 5000 to complete.

A visualization of the massive backlog of 9.5 million compound-disease combinations funneling through a narrow annual trial capacity, illustrating why the process will take until the year 5000 to complete.

Tier 2: Combination Therapies (Still Defensible)

Modern medicine increasingly relies on multi-drug regimens. Oncology, HIV treatment, cardiology, and psychiatry all routinely use drug combinations. Testing pairwise combinations of safe compounds:

45.1B combinations combinations ÷ 3.30k trials/year (95% CI: 2.64k trials/year-3.96k trials/year) trials/year = 13.7M years (95% CI: 11.6M years-16.3M years)

That’s longer than Homo sapiens has existed as a species. Using only compounds with established safety data. Testing only pairs, not triplets or higher-order combinations.

A comparison illustrating the vast gap between current clinical trial capacity and the millions of years required to test all possible dual-drug combinations.

A comparison illustrating the vast gap between current clinical trial capacity and the millions of years required to test all possible dual-drug combinations.

Tier 3: The Full Chemical Universe (Illustrative)

Beyond known compounds lies the vast chemical space of 10^23 to 10^60 drug-like molecules. At current trial capacity, exploring even the synthesizable fraction would take longer than the universe has existed.

A scale comparison showing the tiny dot of known compounds against the massive, expansive sphere of the theoretical chemical universe (10^60 molecules).

A scale comparison showing the tiny dot of known compounds against the massive, expansive sphere of the theoretical chemical universe (10^60 molecules).

We haven’t even started.

“But Won’t We Run Out of Easy Discoveries?”

Some will argue “diminishing returns”: the more we search, the harder each new discovery becomes. This assumes we’ve already picked the low-hanging fruit.

A visualization of the vast gap between current therapeutic exploration (less than 1 percent) and the theoretical 37 percent threshold where diminishing returns begin to outweigh the benefits of compound learning.

A visualization of the vast gap between current therapeutic exploration (less than 1 percent) and the theoretical 37 percent threshold where diminishing returns begin to outweigh the benefits of compound learning.

We haven’t picked any fruit. We’re still in the parking lot.

Diminishing returns only dominates after you’ve explored a significant fraction of the space. At <1% explored, every trial teaches us something new about human biology, improving our ability to predict which combinations will work next. Learning effects compound; depletion effects are negligible.

The math: even with weak learning, diminishing returns only matters after exploring ~37% of therapeutic space. For combination therapies, that crossover is ~7 million years away at current pace.

You cannot have diminishing returns when you’ve barely started looking. See the formal analysis for the equations.

The Fix

The Oxford RECOVERY Trial (COVID-19) proved that large-scale pragmatic trials can run for ~$500 (95% CI: $400-$2.50K) per patient and deliver results in <100 days142.

$500 (95% CI: $400-$2.50K) vs $41K (95% CI: $20K-$120K).

With a decentralized framework for drug assessment, trial capacity increases by 12.3x (95% CI: 4.19x-61.3x):

9.50M combinations combinations ÷ 40.6k trials/year (95% CI: 16.3k trials/year-170k trials/year) trials/year = 234 years (95% CI: 56 years-584 years)

Still a long time, but within the planning horizon of institutions. The first few decades would capture the highest-value discoveries, the low-hanging fruit that current trial scarcity forces us to ignore.

More importantly, dFDA builds permanent infrastructure for systematic therapeutic exploration. The goal isn’t to “finish”, it’s to stop leaving cures on the table while we form subcommittees.

Side-by-side comparison of costs and efficiency gains using decentralized pragmatic trials versus traditional methods.

Side-by-side comparison of costs and efficiency gains using decentralized pragmatic trials versus traditional methods.

The Bottom Line

Humanity has empirically tested less than 1% of the drug-disease relationships that can be tested using existing safe compounds.

If you include the larger chemical universe, humanity has explored roughly one-quadrillionth to one-quintillionth of what is possible (10^-12 to 10^-50).

The overwhelming majority of therapeutic potential remains untouched. Not because cures are impossible. Because trials are too expensive and too slow.

A decentralized, automated pragmatic trial system is the only route to systematically explore the remaining 99.999…% of medical space.

You’re not in the parking lot because there’s nothing inside Disney World. You’re in the parking lot because the entrance fee is $100 million and requires 17 years..

A visualization contrasting the tiny fraction of explored medical relationships against the vast, untapped chemical universe, separated by the barriers of cost and time.

A visualization contrasting the tiny fraction of explored medical relationships against the vast, untapped chemical universe, separated by the barriers of cost and time.

The rides are inside. The cures are inside. You just need to open the gate.