
The Great Conversation: What Happens When a Generation of Scientists Chooses Another Path?

Hello everyone, and welcome to The Great Conversation.
Scientific discoveries often appear to arrive suddenly.
A new medicine is approved.
A new technology emerges.
A disease becomes treatable.
A telescope reveals something we have never seen before.
A breakthrough changes what we thought was possible.
But breakthroughs rarely begin at the moment we hear about them.
They often begin years, sometimes decades, earlier.
They begin when a student becomes fascinated by a question.
When someone enters a laboratory for the first time.
When a doctoral student spends years learning how to conduct research.
When an experienced scientist mentors someone younger.
When a researcher receives the funding necessary to pursue an idea whose usefulness may not yet be obvious.
Science therefore depends on something that receives far less attention than scientific discoveries themselves.
A pipeline.
A recent New York Times investigation examines what is happening to that pipeline in the United States.
Federal research programs that have long helped finance doctoral students and early-career scientists have experienced grant cancellations, funding delays and significant uncertainty.
The Times analyzed data from 27 leading research universities and found substantial declines in new Ph.D. enrollment.
Between 2024 and 2025, new enrollments declined by about 18 percent in engineering programs.
Science and health disciplines declined by roughly 10 percent.
Mathematics and computer science programs also declined by approximately 10 percent.
Behind those numbers are individual decisions.
Students deciding whether pursuing a doctorate still makes sense.
Young researchers wondering whether an academic career remains viable.
Scientists trying to determine whether they can establish laboratories.
Researchers considering opportunities outside the United States.
And people who may have spent much of their lives preparing for scientific careers beginning to imagine entirely different futures.
The political debate surrounding these changes is significant.
The Trump administration argues that federal science funding should become more accountable, less politicized and more focused on research that advances national priorities.
It has also announced initiatives intended to strengthen American participation in science and technology, including programs designed to encourage U.S. students to pursue advanced degrees in fields such as artificial intelligence, science and engineering.
Critics, including university leaders and researchers, argue that abrupt grant cancellations, funding delays and uncertainty can weaken the scientific system regardless of the intentions behind those changes.
That disagreement deserves serious debate.
Governments should certainly ask how public money is being spent.
Research institutions should be accountable.
Scientific priorities can change.
No funding system should be beyond examination simply because it has existed for a long time.
But beneath the political disagreement lies a deeper question about how scientific capacity is actually created.
Because a laboratory can lose funding quickly.
A doctoral program can reduce admissions quickly.
A researcher can accept another career quickly.
A talented scientist can move to another country relatively quickly.
But producing an experienced scientist is not quick.
It takes years.
Often many years.
And this creates an unusual asymmetry.
Scientific infrastructure can be disrupted much faster than scientific expertise can be created.
Consider what it takes to produce a researcher capable of leading an advanced laboratory.
Years of undergraduate education.
Years of doctoral training.
Mentorship.
Experiments.
Failures.
Publications.
Collaboration.
Specialized knowledge.
Postdoctoral research.
Grant applications.
Experience managing people and projects.
By the time someone becomes capable of independently leading important scientific work, society may already have invested more than a decade in that person's development.
And much of that investment is invisible.
We see the scientist.
We rarely see the ecosystem that produced the scientist.
Universities.
Government agencies.
Laboratories.
Professors.
Research grants.
Fellowships.
Equipment.
Other students.
Institutional knowledge.
International collaboration.
Previous generations of scientists.
Scientific excellence is therefore not simply a collection of intelligent individuals.
It is an ecosystem.
And ecosystems depend heavily on continuity.
That may be why uncertainty can matter almost as much as funding itself.
Imagine being 22 years old and considering a Ph.D.
The decision may require five, six or more years of your life.
The financial rewards are uncertain.
The work can be extraordinarily demanding.
Academic positions are already highly competitive.
Now add uncertainty about whether the programs, fellowships and laboratories supporting your training will continue to exist.
The rational decision may change.
Not because you suddenly stopped loving science.
But because the path itself became harder to trust.
And when enough individuals make that decision, something interesting happens.
A personal decision becomes a national outcome.
One student chooses industry.
Another abandons a doctorate.
Another scientist accepts a position overseas.
Another laboratory admits fewer students.
Another professor retires without someone prepared to continue the work.
Individually, none of these decisions necessarily looks historic.
Collectively, they can reshape a country's scientific capacity.
This is where the issue becomes larger than any particular administration.
Every society inherits institutions built by people who came before it.
Universities.
Scientific agencies.
Research laboratories.
Infrastructure.
Legal systems.
Public institutions.
Knowledge itself.
Each generation therefore faces two responsibilities.
It must decide what should be changed.
And it must decide what must be preserved long enough for something better to replace it.
Those two responsibilities can conflict.
Institutions can become inefficient.
Funding priorities can become outdated.
Bureaucracies can become resistant to change.
Reform may be necessary.
But reforming a system and destabilizing a system are not necessarily the same thing.
The difficulty is knowing the difference while the transformation is happening.
This is especially important in science because the consequences often appear much later.
If a bridge collapses, the failure is immediate.
If a company loses customers, the financial consequences appear relatively quickly.
But if fewer scientists begin training today, the consequences may not become fully visible for ten or twenty years.
The missing discovery has no headline.
The experiment that was never conducted produces no data.
The laboratory that was never created leaves no empty building.
The scientist who chose another career may never know what they might have discovered.
This makes scientific capacity strangely difficult to measure.
We can count the discoveries that happened.
We cannot easily count the discoveries that never had the opportunity to happen.
And history suggests that some of the technologies that eventually transform civilization begin as research whose practical value is not immediately obvious.
Scientific research has contributed to technologies ranging from modern medicine and computing to satellite navigation and communications.
But those outcomes were not always predictable when the underlying research began.
That uncertainty is part of science.
We fund questions partly because we do not yet know where their answers will lead.
And this creates another tension.
Governments naturally want results.
Taxpayers reasonably want accountability.
But discovery often requires allowing intelligent people to investigate questions whose eventual usefulness cannot be guaranteed.
There is no perfect formula for balancing those needs.
Too little accountability can waste resources.
Too much insistence on immediate usefulness can discourage exploration.
Science lives somewhere between those two risks.
And the competition for scientific talent makes the question even larger.
Scientists are unusually mobile.
Knowledge crosses borders.
Researchers collaborate internationally.
Countries compete to attract talented students and experienced scientists.
For much of modern history, the United States benefited enormously from this movement.
People came from around the world to study, conduct research and build careers in American universities, laboratories and companies.
That means scientific leadership is never simply possessed.
It must continually be renewed.
A country can inherit scientific institutions.
It cannot inherit the next generation of scientists automatically.
Those people still have to choose the path.
They have to believe the years of training are worthwhile.
They have to believe opportunities will exist when the training is finished.
They have to believe there will be somewhere to conduct the work.
And perhaps most importantly, they have to believe that the society around them still values what they are preparing themselves to do.
This does not mean every research program must survive.
It does not mean every grant deserves funding.
It does not mean science should be protected from political oversight.
But it does mean that decisions affecting scientific institutions operate on a longer clock than ordinary politics.
Presidential terms last years.
Congressional majorities change.
Budgets change annually.
Scientific generations unfold across decades.
And the consequences of decisions made during one political moment may still be developing long after the people who made those decisions have left office.
Perhaps that is the larger lesson.
Civilizations are constantly borrowing from their own future.
Education is a bet on people we hope will become capable later.
Infrastructure is a bet that people will need it later.
Basic research is a bet that knowledge we cannot immediately use may eventually become indispensable.
Training scientists is another such bet.
We invest in people today because we believe someone tomorrow will need the questions they learn how to ask.
And sometimes the greatest value of that investment will not become visible until long after the original decision has been forgotten.
Which raises a deeper question...
When the knowledge and discoveries a society will need twenty years from now depend on the people it chooses to invest in today, how should we decide what can safely be cut, reformed or disrupted without accidentally weakening a future we cannot yet see?
Take a moment to reflect.
And click the link below to see more on this subject.

