I heard a proverb: The fact is silent until we have a question. (Пока нет вопроса, факт молчит).
Until we have a hypothesis to test, the fact is not evidence for anything in Bayesian sense.
But we need to make a step above a single hypothesis - from one hypothesis to research program. Here all facts become relevant.
1) Before going into historical penicillin details, let's think how transferrable is the penicillin example to our world?
Arguably only to a restricted extent.
Most importantly, the biochemical R&D in 1938 was about three orders of magnitude less than in 2026. Inflation adjusted NIH funding was a few thousand times less than now (and btw Rockefeller funding was among the greatest sources of funding globally comparable to NIH). There were ~300-1000x less biochemical researchers in the world compared with 2026. The role of individual contributors has markedly decreased since then, the science is increasingly institutional. Unless it's cryonics or some another field with low number of scientists and funding.
2) "Could we make penicillin earlier?" - it depends on what we count as scenarios here.
a) Under realistic conditions probably no. Before the war created an enormous centralized and urgent demand, penicillin was too expensive to scale. One might make the mice experiment in 1935 and it would still stay too expensive until the war.
b) If we imagine a vitalist US president at the time - surely yes.
3) Penicillin was known among experts in 30s but its research was viewed as high-risk high-reward pursuit, meaning that we should adjust the contribution to the survivorship bias. Some high-risk high-reward pursuits give large results (hence high-reward), but high-risk means the apriori expectation is significantly lower than it appears aposteriori for the winners.
4) Would it be still discovered and scaled during the war if Florey/Chain/Heatley worked on smth else? Experts knew penicilline at this time as a uniquely potent substance against medically (and military-medically) important pathogens (high reward) but notoriously difficult and expensive to purify and produce. The war created huge well-paid market and incentive for research.
5) generally agree regarding Fleming - it's clear from the story that whether he discovered penicillin a few years earlier or later didn't play a significant difference.
6) Have they accelerated LEV? By accelerating penicillin itself - probably no; it's still bottlenecked by other things, including the IT revolution for data storage and analysis.
However the strongest argument for acceleration would mention the influence of penicillin scaling success story on post-war NIH funding increase (10x in 8 years of Eisenhower presidency). This influence is real, so they can have some claim.
Before Fleming
The history of antibiotics began long before Fleming, although popular culture tends to attribute almost the entire breakthrough to him.
Half a century before Fleming, scientists had already observed microbial antagonism and the therapeutic potential of mold several times. But each observation looked like an isolated biological curiosity rather than the beginning of a new technological platform.
Around Fleming
In 1922, Fleming discovered lysozyme, a substance found in tears, saliva, and mucus that destroys certain bacteria.
In 1928, he noticed that mold had accidentally contaminated one of his Petri dishes and killed the bacteria around it. In 1929, he published the results and named the active substance penicillin.
Fleming understood that the substance might potentially be used against infections. But the practical emphasis of his paper was much more modest: penicillin could become a useful laboratory tool and perhaps a topical antiseptic. The path toward a systemic drug remained unclear.
In 1930, Fleming’s former student Cecil Paine used crude mold filtrate to treat eye infections. The effect already worked in patients, although only locally. But even this did not launch a major development program. Penicillin remained unstable, difficult to purify, and seemingly impossible to scale.
For almost a decade, the discovery had no owner.
Systematic search and manufacturing
In late 1938, Howard Florey and Ernst Chain at Oxford decided to expand their research from lysozyme to naturally occurring antibacterial substances as a broader class. They were not initially developing one specific drug. They were systematically searching for promising effects.
While reviewing older papers, Chain came across Fleming’s nearly forgotten 1929 publication. Penicillin became one of the candidates worth testing.
In May 1940, the team of Florey, Chain, and Norman Heatley showed that penicillin could save infected mice. Eleven years after Fleming’s paper, it became clear that the substance worked inside a living organism.
In February 1941, penicillin was administered systemically to a critically ill patient for the first time. He rapidly improved, but the supply ran out and he died. The bottleneck was no longer biology. It was manufacturing.
In 1941, Florey and Heatley traveled to the United States. The Department of Agriculture laboratory in Peoria improved the growth medium and increased penicillin yields roughly tenfold. The government and pharmaceutical companies then developed deep-tank fermentation and built industrial-scale production facilities.
Production in the United States rose from 21 billion units in 1943 to 1,663 billion units in 1944. By the time of the Normandy landings, enough penicillin was available to treat severely wounded soldiers.
Conclusion
For roughly 60 years, humanity circled around the idea of antibiotics without recognizing a technological platform in the repeated observations. Even Fleming identified an especially important signal, yet his paper also remained largely unnoticed.
The revolution happened when a team began systematically screening candidates on the basis of accumulated knowledge.
It is worth asking how many breakthroughs remain unnoticed around us simply because no team is conducting a systematic search within a specific field.