When the United States detonated the first atomic bomb in July 1945 — the Trinity test, weeks before Hiroshima — only one country had nuclear weapons. By 1949 the Soviets had broken the monopoly. The United Kingdom (1952), France (1960), China (1964), India (1974's "peaceful" test, weaponized later), Israel (probably from the late 1960s, never officially acknowledged), Pakistan (1998), and North Korea (2006) followed. Iran has, at various points, been months to a few years from a device. About a dozen other states — Japan, South Korea, Germany, Saudi Arabia, Brazil among them — could plausibly build weapons within a year if they chose to, since the hard part is enrichment, not design. The N is not stable. Every additional nuclear state changes the calculations of all the others.
The Nuclear Non-Proliferation Treaty (NPT) of 1968 is the single most successful piece of arms-control legislation in history — much of the reason there are nine nuclear states rather than the thirty that President Kennedy feared by the 1970s. Its grand bargain had three parts: non-nuclear states would forgo weapons; existing nuclear powers would eventually disarm under Article VI; and in return all signatories would get access to civilian nuclear technology. The deal has eroded. The recognized nuclear states have not disarmed in any meaningful sense — they are modernizing instead — and the non-nuclear states have noticed. The example of Ukraine, which transferred its inherited Soviet-era arsenal (then the world's third-largest) to Russia in 1994 under the Budapest Memorandum in exchange for security assurances that did not survive 2014, is now read in capitals from Riyadh to Seoul to Warsaw as a cautionary tale about giving up the bomb. Libya's Gaddafi, who surrendered his weapons program in 2003 and was overthrown eight years later, only sharpened the lesson. The risk of further proliferation is therefore not abstract; it is being calculated, country by country, against the backdrop of receding American security guarantees and an enrichment technology that gets cheaper every decade.
The nuclear non-proliferation regime is plausibly the most under-priced strategic risk on the planet. A world of fifteen or twenty nuclear states would be qualitatively more dangerous than the current one — more crises, more chances for miscalculation, shorter decision timelines, and more combinations in which deterrence might fail because each new arsenal interacts with every other. The last bilateral U.S.–Russia treaty, New START, expires in 2026 with no successor in sight, just as China races to expand its own stockpile. The chance of a nuclear weapon being used in anger somewhere in the next twenty years is not zero, and on present trends it is going up, not down.