Scientific Sovereignty Versus Nuclear Proliferation Risks

00:00:10 I'm Mahmoud Hessabi, an Iranian physicist and educator. In this imagined

00:00:16 educational debate, we're bringing our perspectives to questions that reach

00:00:21 beyond our lifetimes. Professor Einstein, a country can buy a sophisticated

00:00:27 machine and still have to ask permission to understand it. That is the problem I

00:00:33 want to put on the table. Peaceful nuclear access should include training

00:00:38 people, building laboratories, checking safety and choosing useful applications.

00:00:44 For developing countries, that freedom matters: scientific independence cannot

00:00:50 rest on whatever knowledge a supplier happens to offer. Yes, nuclear technology

00:00:56 brings unusual dangers. The right response is to distinguish activities and

00:01:01 govern them properly. A medical laboratory and a sensitive fuel facility require

00:01:07 different controls. I support that distinction. But the starting presumption

00:01:13 should be participation. A country's scientists should not need their nation to

00:01:19 become wealthy and powerful before they are trusted to contribute.

00:01:24 I'm Albert Einstein, a physicist with a stubborn interest in what happens after

00:01:29 an elegant idea leaves the laboratory. Dr. Hessabi, I support the participation

00:01:35 you describe. But our motion makes a further claim: that the benefits outweigh

00:01:40 the geopolitical risks. A right deserves protection; a prediction still needs

00:01:45 evidence. Some nuclear capabilities create consequences far beyond the

00:01:50 institution operating them. Neighbours react, governments change, and technical

00:01:56 capacity survives the promises made at the opening ceremony. Nor does scientific

00:02:01 independence require a country to develop every prestigious field. A nation can

00:02:07 make brilliant contributions through medicine, computing or materials science

00:02:12 while deciding that a particular nuclear investment is a poor choice. So give

00:02:17 scientists access. Give citizens choices. But don't let the word 'sovereignty'

00:02:23 settle the balance before we have examined the project,

00:02:26 the alternatives and the people who will carry its risks.

00:02:30 You are right that a country need not pursue every field. But choosing freely

00:02:36 requires more than being handed a shopping catalogue. Suppose a university can

00:02:42 purchase equipment but cannot train the specialists who maintain it, investigate

00:02:49 a defect or develop the next generation. It has acquired a service. Its capacity

00:02:55 to question the supplier remains weak. Peaceful nuclear cooperation can help

00:03:01 build demanding institutions: measurement laboratories, technical education,

00:03:07 radiation protection and disciplined quality control. Those capabilities can

00:03:13 serve work beyond a single installation. I am not promising that a reactor

00:03:18 automatically creates a scientific culture. It plainly does not. My argument

00:03:24 is that responsible access lets a country build that culture on its own terms.

00:03:30 Let an independent assessment reject a bad project. What I oppose is rejecting a

00:03:37 country's participation before its scientists have even presented one.

00:03:42 And I oppose that exclusion too. But notice where the argument has moved: from

00:03:47 a right to participate to a claim about which investment builds competence best.

00:03:53 A demanding project can strengthen institutions. It can also consume the

00:03:58 people and money those institutions needed elsewhere. Training an excellent

00:04:03 nuclear engineer does not tell us whether that was the best use of the next

00:04:08 scholarship. The comparison matters. What would the same commitment achieve in

00:04:13 grid reliability, public health, laboratories or technical colleges? Count

00:04:18 the benefits of nuclear cooperation, certainly. Also count the benefits of the

00:04:24 opportunities it displaces. I would gladly fund independent scientific

00:04:29 education. I would not attach a nuclear programme to it as a certificate of

00:04:34 national maturity. Scientific confidence includes being able to say, 'We

00:04:39 understand this technology perfectly well, and we have other priorities.'

00:04:44 Then let's count the cost of dependence with the same enthusiasm. Imagine a

00:04:49 competent local team facing a supplier that stops providing parts or technical

00:04:55 support during a political dispute. The facility still belongs to the country.

00:05:00 Its practical choices have suddenly narrowed. Buying from abroad can be

00:05:06 sensible. Making the entire system depend on one relationship is a different

00:05:11 proposition. Independent expertise helps a country challenge a contract, inspect

00:05:17 a repair and judge whether another supplier offers a workable alternative.

00:05:22 You emphasise capabilities that outlast political promises. I emphasise

00:05:27 obligations that may not survive political pressure. My preferred programme

00:05:33 would develop local competence, diversify external partnerships and reserve the

00:05:38 most sensitive activities for much stronger scrutiny. That gives us a way to

00:05:44 reduce dependence without demanding national ownership of everything. Why

00:05:49 should that version of scientific independence

00:05:53 begin with a presumption against it?

00:05:55 It shouldn't. That is a sensible version, and I accept it. But supplier

00:06:00 dependence and sensitive technical capacity do not have the same

00:06:04 reversibility. A broken supply relationship may be repairable through

00:06:09 another contract, another partner or a different design, although the transition

00:06:14 can be painful. Once sensitive expertise and infrastructure exist, cancelling a

00:06:19 licence does not make them disappear. That difference matters to neighbours.

00:06:24 They cannot simply read today's promise and stop worrying about tomorrow's

00:06:29 leadership. So strengthen local competence where it makes the system safer and

00:06:34 more resilient. Don't assume that each additional layer of domestic capability

00:06:39 adds independence without adding a different exposure. The useful question is

00:06:45 specific: which capability, under whose control, with what benefit and what

00:06:50 consequence if cooperation collapses? Your account becomes stronger when it

00:06:55 answers those questions. It becomes weaker when

00:06:58 independence supplies the answer in advance.

00:07:01 Agreed. And 'under whose control' must include the citizens of the country, not

00:07:07 just its ministers. A sovereign right belongs to a political community. It is

00:07:13 not a personal privilege of whoever signs the contract. Communities near a

00:07:19 proposed facility should receive intelligible information, independent

00:07:24 advice and a real chance to influence the decision. Publish the financing

00:07:30 assumptions. Protect the engineer who reports a fault. Give the regulator

00:07:36 authority to stop operations. Let people challenge the plan through institutions

00:07:42 that can actually respond. Those are demanding conditions. They are also

00:07:47 reasons to build accountable capacity, rather than leave every important

00:07:53 judgement inside a foreign supplier's office. I want a country capable of saying

00:07:59 yes responsibly and no effectively. A government that cannot tolerate its own

00:08:05 scientists asking awkward questions has no business

00:08:09 presenting its programme as scientific independence.

00:08:14 That is the strongest definition of sovereignty you've offered. But it also

00:08:19 exposes a serious problem: a national prestige project can make the right to say

00:08:25 no politically expensive. An accountant questions the cost and is called

00:08:30 defeatist. An engineer questions the timetable and is told to stop embarrassing

00:08:35 the country. A nearby community raises concerns and is accused of obstructing

00:08:41 progress. Those are possibilities, not accusations against every developing

00:08:46 state. Wealthy countries can behave that way too. A paper right to appeal will

00:08:52 not suffice if careers depend on keeping the project alive. We need financing

00:08:57 and institutions that can survive cancellation, not merely institutions

00:09:02 authorised to discuss it. Can your programme still be judged a national

00:09:07 success if an independent regulator stops it before it opens? If the honest

00:09:13 answer is no, political symbolism has already

00:09:16 taken control of scientific judgement.

00:09:19 Yes. Stopping a project for a sound reason can be a success. It means the

00:09:24 country has learned something before committing more people and money. That is

00:09:29 why I would begin with useful, manageable capabilities and make larger

00:09:34 commitments conditional on demonstrated need and institutional readiness.

00:09:39 Education, laboratories and regulatory competence should have value even if a

00:09:44 proposed power station is never built. Financing should expose long-term

00:09:49 obligations instead of hiding them behind a construction headline. Funds for

00:09:54 safety and eventual decommissioning should not depend on future politicians

00:09:59 feeling generous. There will still be difficult choices. But compare that design

00:10:05 with a turnkey promise whose real costs are understood only by the vendor.

00:10:10 Independence is the ability to interrogate a proposal before becoming trapped

00:10:15 inside it. The country that can cancel intelligently has gained something

00:10:20 substantial, even without a ribbon-cutting ceremony.

00:10:24 Before the large commitment, yes. What about after it? A partly completed

00:10:28 project changes the politics. Jobs depend on continuation. Loans remain due.

00:10:33 Leaders have promised an opening date. The argument becomes, 'We have spent too

00:10:39 much to stop,' even when the remaining expenditure no longer makes sense. Your

00:10:44 milestones help only if failure at a milestone can actually halt the next

00:10:49 payment. Decommissioning funds help only if they are protected, sufficient and

00:10:54 honestly assessed. And there is a generation that never attends the

00:10:59 consultation: the people who inherit the site, the waste and the institutions

00:11:04 responsible for them. I am not using those obligations to rule out nuclear

00:11:09 power. I am asking us to include them in the case for it. The opening ceremony

00:11:14 lasts an afternoon. Responsible ownership has

00:11:17 to survive the people standing on the stage.

00:11:20 Let's bring that responsibility into a place where delay also has a human cost:

00:11:26 a hospital. Reliable nuclear medicine depends on more than buying a finished

00:11:31 product. It needs trained staff, suitable facilities, quality control and

00:11:37 dependable supply. Different radioisotopes can come through reactors,

00:11:42 accelerators or generators; the right arrangement depends on the application. A

00:11:48 country should be able to develop the expertise to choose and operate that

00:11:53 arrangement. Some production may be local, some regional, some imported. The

00:11:59 objective is reliable care, not a flag over every component. This is why I

00:12:04 resist treating nuclear access as one enormous decision about one enormous

00:12:10 plant. It is a field of decisions, many with benefits that ordinary people can

00:12:15 recognise immediately. We should compare these arrangements seriously. But

00:12:21 denying the necessary skills because another part of nuclear technology is

00:12:26 sensitive would be extraordinarily poor judgement.

00:12:30 On that example, we agree. A dependable medical service is a strong case for

00:12:35 cooperation, and often for local capability too. But the example works

00:12:40 because you have identified a particular need and matched the technology to it.

00:12:46 The benefit is concrete. The delivery options can be compared. Safety

00:12:51 obligations remain visible. What it does not do is establish that the whole

00:12:56 policy category has a favourable geopolitical balance. A country could make

00:13:01 excellent medical investments and still make an extremely unwise decision

00:13:06 elsewhere in its nuclear programme. I would protect access to medical training

00:13:12 and suitable technology while judging sensitive fuel activities separately. That

00:13:18 is a meaningful distinction, not a loophole for excluding poorer countries. A

00:13:23 hospital gives us a reason to support the hospital's needs. It should not have

00:13:29 to carry the argument for every other ambition attached to the word 'nuclear'.

00:13:34 Fair. Then consider electricity on its own merits. A growing economy needs

00:13:40 reliable power with lower emissions. Nuclear generation can contribute to that

00:13:46 objective. Whether it makes sense depends on the grid, financing, construction

00:13:52 capability, other resources and the country's long-term plan. A responsible

00:13:57 government should compare it with renewables, storage, transmission,

00:14:02 efficiency and other available choices. It should not be told that one option

00:14:08 belongs to a different class of nations. Nor should a useful partnership force a

00:14:14 country to outsource all the judgement that makes the partnership safe. Local

00:14:20 people need the ability to understand the plant, scrutinise performance and hold

00:14:26 operators accountable. That is my claim about access: let the country develop

00:14:32 enough competence to choose well, and retain enough authority to govern the

00:14:38 choice. The possibility of a bad investment justifies a rigorous

00:14:42 comparison. It does not justify closing the subject.

00:14:46 The subject should stay open. The comparison should stay open too, including

00:14:52 after political leaders have found their favourite answer. Reliable low-carbon

00:14:57 electricity is a benefit. How soon it arrives, what it costs and what the grid

00:15:02 can accommodate determine how valuable that benefit is in a particular system. A

00:15:07 technically impressive project can leave an urgent problem unresolved while

00:15:13 money and attention are tied up. Another country may have the institutions,

00:15:18 demand and financing that make nuclear power a sound choice. I expect the

00:15:22 answers to differ. That is where I still resist the motion's general judgement.

00:15:28 Scientific independence gives a country the capacity to investigate. It does not

00:15:33 make the result of that investigation predictable. I am perfectly willing to

00:15:38 support a strong nuclear project. I want its case to remain strong when we

00:15:43 remove national prestige from the calculation.

00:15:46 Then address the geopolitical risk directly, and apply the controls where

00:15:52 that risk actually arises. I would require declared activities, reliable

00:15:58 material accounting, independent domestic oversight and international

00:16:03 verification. Stronger access for inspectors can improve their ability to

00:16:09 identify undeclared activity. It cannot promise perfect knowledge. For more

00:16:15 sensitive proposals, I would demand more: a demonstrated civilian need, a

00:16:20 credible regional security assessment and limits enforceable through the

00:16:26 cooperation agreement. Some proposals would fail that test. They should fail it.

00:16:32 My defence of access has room for refusal. It does not have room for a permanent

00:16:38 assumption that nationality predicts misconduct. A system that expands useful

00:16:44 cooperation and constrains dangerous choices can deliver a better balance than

00:16:50 either blind trust or broad exclusion. You have shown why the risks deserve

00:16:56 weight. You have not shown that they should define the entire field's future.

00:17:02 I don't want them to define the entire field. I want their consequences to

00:17:07 receive more than an entry in the inspection budget. A sensitive capability

00:17:12 may alarm a neighbour even without a proven violation. That neighbour may

00:17:17 respond with its own programme, coercive pressure or a military threat. Each

00:17:22 side can describe its conduct as defensive while making the other less secure.

00:17:27 Better verification can reduce uncertainty. It cannot force leaders to

00:17:31 trust one another, and it cannot prevent every political rupture. So the balance

00:17:36 is not simply civilian benefits minus the probability that inspectors miss

00:17:41 something. It also includes what the capability does to relationships between

00:17:46 states. Your conditions are useful. They make a proposal more defensible. But a

00:17:51 proposal can satisfy many technical conditions and still enter a regional

00:17:56 setting in which its political effects are dangerously difficult to control.

00:18:01 Then cooperation must change those relationships as well as inspect

00:18:06 equipment. Imagine a shared regional institution with meaningful participation

00:18:11 by its members, transparent decisions and independent review. It could support

00:18:17 training, supply resilience and agreed civilian services while limiting

00:18:22 unilateral control over sensitive activities. I am proposing a design

00:18:26 principle, not pretending that every region is ready for it. Where trust is too

00:18:32 weak, start with narrower cooperation. Where participation is merely ceremonial,

00:18:38 call it what it is and improve the agreement. The point is that sovereignty

00:18:43 need not mean isolation. A country can exercise agency through institutions it

00:18:48 helps govern. If we offer only dependence on a supplier or a wholly national

00:18:54 programme, we manufacture an unnecessarily harsh choice. Shared authority gives

00:18:59 us another possibility. Does your position leave room for building that

00:19:04 possibility, or must distrust always get the first and last word?

00:19:09 It leaves plenty of room. But a shared institution needs a plan for the day its

00:19:14 members stop sharing a political outlook. Who keeps access open if the host

00:19:19 government closes the gate? Who pays when a member leaves? Which obligations

00:19:25 survive withdrawal? Can a participant challenge a decision without losing the

00:19:30 service its hospitals or electricity system rely on? Joint ownership helps only

00:19:35 if those answers are credible. It is not a substance we sprinkle over a facility

00:19:41 to make distrust evaporate. I would begin with arrangements whose failure is

00:19:46 manageable, and expand them only when experience supports expansion. Even then,

00:19:51 I would distinguish common training and supply services from activities carrying

00:19:56 greater proliferation risk. I support building trust through cooperation. I

00:20:02 oppose spending tomorrow's trust before today's

00:20:05 institutions have shown that they can hold it.

00:20:08 And powerful states must help make those institutions credible. They cannot

00:20:13 demand restraint while treating dependable access as a favour they may withdraw

00:20:19 whenever convenient. A fair agreement needs supply terms that are clear in

00:20:25 advance, a process for resolving disputes and restrictions tied to specified

00:20:30 conduct. Essential civilian cooperation should not become a casual instrument of

00:20:36 pressure. The responsibilities of existing nuclear-weapon states matter too.

00:20:42 Demands for restraint gain credibility when those states take their own

00:20:47 disarmament commitments seriously. None of that excuses secret activity or gives

00:20:53 another country permission to build weapons. It addresses whether the rules

00:20:59 deserve confidence. You ask developing countries to trust institutions rather

00:21:04 than maximise national control. I am asking the states with the greatest

00:21:10 influence over those institutions to make that trust rational. The bargain has

00:21:16 obligations on both sides. Otherwise it is

00:21:19 simply a lecture with a contract attached.

00:21:22 Yes. A lecture is not a bargain, and selective responsibility undermines the

00:21:27 rules it claims to defend. I accept your demand for dependable civilian

00:21:31 cooperation and serious action by existing weapon states. They cannot reserve

00:21:37 uncertainty for everyone else while describing their own power as a guarantee

00:21:42 of stability. But justice still requires us to choose remedies that reduce

00:21:47 danger. Giving every country an equally risky option would make the distribution

00:21:52 more equal without making the world safer. So let us strengthen access, dispute

00:21:57 resolution and reciprocal restraint together. Where a supply arrangement

00:22:02 fails, repair it and hold the responsible parties accountable. Don't assume its

00:22:07 failure establishes the case for every form of national replacement. The

00:22:12 strongest part of your argument is that unfair rules weaken cooperation. The

00:22:17 next step is better rules and fewer dangerous dependencies,

00:22:21 with the risks of the replacement counted honestly.

00:22:24 Then here is the policy I am defending. Start with a protected right to peaceful

00:22:30 participation: education, research, appropriate applications and the knowledge

00:22:36 needed for independent oversight. Require transparent civilian purposes and

00:22:41 obligations proportionate to the activity. Make major projects earn their

00:22:47 financing through a comparison with alternatives, public scrutiny and a

00:22:52 credible plan for long-term responsibilities. Give sensitive

00:22:56 activities a separate, much tougher review. No automatic entitlement follows

00:23:02 from calling something national development. Make cooperation dependable

00:23:07 enough that countries can choose shared provision without surrendering their

00:23:13 ability to question it. On that basis, I judge the benefits of peaceful access

00:23:18 to outweigh its geopolitical risks. Broad scientific participation offers

00:23:24 recurring public value; concentrated risks demand concentrated controls. I have

00:23:29 not withdrawn the motion. I have explained how to honour a right responsibly. A

00:23:35 right without conditions would be reckless. Conditions

00:23:39 that hollow out the right would be dishonest.

00:23:42 You have built a serious policy. Our disagreement is now precise: I support

00:23:48 much of the policy without accepting the general conclusion you attach to it.

00:23:53 The controls do the difficult work. They decide which activities proceed, where

00:23:59 they proceed and whether the remaining risks are acceptable. In some settings,

00:24:05 they may rule out the most consequential projects entirely. That is not a demand

00:24:11 that you prove every project succeeds. A general judgement can tolerate

00:24:16 exceptions. But it still needs an account of why benefits prevail across the

00:24:21 varied circumstances covered by the motion, including political failure after

00:24:27 an initially responsible decision. You have established a strong right to be

00:24:32 considered fairly. You have established substantial benefits in particular

00:24:38 applications. You have not established that nuclear access is essential to

00:24:43 scientific independence or that its geopolitical balance

00:24:47 can be presumed favourable across developing countries.

00:24:51 The word 'essential' concerns the freedom to choose. A country cannot be

00:24:57 scientifically independent if an entire field is closed to it by someone else's

00:25:03 permanent decision. It need not build everything. It must be able to

00:25:08 investigate, participate and develop what it can justify. Our disagreement is

00:25:13 not about whether every project works. It is about the presumption governing

00:25:19 access. I favour responsible participation because useful knowledge, resilient

00:25:25 institutions and civilian applications have value beyond any single programme. I

00:25:31 accept demanding controls, shared authority where appropriate and the

00:25:36 rejection of proposals whose risks cannot be justified. Those limits make the

00:25:42 right workable. Professor Einstein has made my conditions sharper. He has not

00:25:47 persuaded me that a country's proper place in science can be decided from

00:25:53 outside it. Let countries earn trust through conduct. Give their scientists a

00:25:59 fair chance to turn that trust into public benefit.

00:26:02 And I favour a fair chance too. No country should be assigned a permanent place

00:26:08 at the bottom of the scientific table. But access, ownership and a judgement of

00:26:13 net benefit remain different questions. Protect the first. Scrutinise the

00:26:19 second. Earn the third through evidence. Dr. Hessabi's strongest contribution

00:26:24 was to make sovereignty answerable to citizens, including their right to stop a

00:26:29 project. My concern is what remains when those protections weaken: obligations

00:26:35 to future generations, capacity that cannot simply be recalled, and neighbours

00:26:40 making their own fearful calculations. None of that erases the benefits. It

00:26:46 prevents us from treating a legitimate aspiration as a reliable forecast. Build

00:26:51 the hospital capability. Support the laboratories. Approve the power project

00:26:56 when its case holds up. Decline the dangerous commitment when it does not. A

00:27:02 scientifically independent country should possess

00:27:05 enough confidence to make every one of those decisions.

00:27:09 I'm J. Robert Oppenheimer, a physicist and former director of the wartime Los

00:27:14 Alamos laboratory. My task here is to judge the argument, including what

00:27:18 survives after a government changes its mind. Hessabi made sovereignty concrete.

00:27:23 His independent regulator, informed citizens and ability to cancel a project

00:27:28 were substantial answers to the danger of national prestige overriding

00:27:32 judgement. He also distinguished useful competence from automatic ownership of

00:27:37 sensitive facilities. Einstein's stronger answer was that even good initial

00:27:42 governance does not settle the later geopolitical effects. Technical capacity,

00:27:47 financial obligations and regional mistrust may outlast the agreement that

00:27:51 made a project acceptable. I am not asking Hessabi to prove that every

00:27:56 application succeeds. The motion permits exceptions. But a general balance still

00:28:01 needs support, and the clearest benefits presented concerned selected

00:28:05 applications. The broader risks remained strongly dependent on context. My vote

00:28:10 goes to Albert Einstein. He defended access while explaining why the net

00:28:15 judgement must remain open, especially where commitments are hardest to reverse.

00:28:20 I'm Marie Curie, a physicist and chemist whose work centred on radioactivity. I

00:28:25 judge this debate by whether its distinctions help scientific work serve

00:28:29 people. Einstein was right to insist on alternatives, institutional competence

00:28:34 and the cost of commitments passed to later generations. Hessabi could not

00:28:39 answer those concerns simply by pointing to a hospital, and eventually he did

00:28:43 much more than that. His final case treated access as a practical freedom:

00:28:48 learning, choosing suitable applications, scrutinising suppliers and accepting

00:28:53 different requirements for different risks. He explicitly accepted that some

00:28:58 proposals should be rejected. That is compatible with a general right, not a

00:29:03 retreat from one. I interpret the motion as a choice about responsible peaceful

00:29:08 participation, not a recommendation that every developing country build a

00:29:12 national nuclear industry. Under that interpretation, Hessabi made the stronger

00:29:17 case that the public benefits justify a presumption of access with enforceable

00:29:22 obligations. My vote goes to Mahmoud Hessabi. Einstein supplied indispensable

00:29:27 cautions, but Hessabi gave those cautions a

00:29:29 workable place inside the right he defended.

00:29:32 I'm Kofi Annan, a former United Nations secretary-general. I will focus on the

00:29:38 difference between an agreement worth building and a claim already established.

00:29:43 Hessabi was strongest on reciprocity. If powerful states expect others to rely

00:29:49 on shared institutions, they must make access dependable, resolve disputes

00:29:54 fairly and take their own responsibilities seriously. His account of sovereignty

00:30:00 also included the public's ability to question its government. That mattered.

00:30:05 Einstein accepted those obligations. His objection was narrower and more

00:30:10 persuasive: an equitable right to participate does not, by itself, establish

00:30:16 that the resulting geopolitical balance is favourable. Host-state control,

00:30:21 institutional failure and regional reactions can alter that balance even when

00:30:26 a programme began responsibly. The motion does not say every project must

00:30:32 succeed. Nevertheless, it links a legitimate right to a broad claim of

00:30:37 necessity and net benefit. I find the right well defended, but the further

00:30:42 claims insufficiently established. My vote goes to Albert Einstein. With

00:30:47 Oppenheimer and me voting for Einstein, and Curie

00:30:50 for Hessabi, the decision is two votes to one.