The fear of a nuclear fire that would consume Earth

In the early years of nuclear research, some scientists feared breaking open atoms might start a chain reaction that would destroy Earth.

A scene in Christopher Nolan's Oppenheimer hinges around the worry some Manhattan Project scientists felt that the first atomic bomb test would ignite Earth's atmosphere.

Edward Teller, so the story goes, first raised the possibility in 1942. It has been debated how seriously this was taken, but it animated the scientists enough to crunch the numbers, concluding planetary immolation was "unlikely".

Pyrotechnic fears of this sort, however, predate serious efforts to make nuclear bombs by decades, going back to a time when cracking atoms seemed, to many, pure science fiction. Though analogies between artificial intelligence and nuclear weapons can be overstretched, the way this concern went from idle speculation to serious concern can teach us lessons regarding fears swirling around today's emerging technologies.

Long ago, the Roman naturalist Pliny the Elder remarked that, given all the combustible matter, it's a daily miracle the globe is not alight. But for the longest time everyone assumed that – though objects may burn – the atoms they are made from are stable, sturdy, and indestructible. When, as the 1900s opened, Marie Curie revealed radioactivity's secrets to the world, this changed overnight.

Curie's discoveries unexpectedly demonstrated a titanic reservoir of energy lurks, locked up, within matter's basement. As they disintegrate, radioelements are unspooling this internal wealth. But all atoms, it was reasoned, contain such magnitudes. These revelations, African American scientist C H Turner remarked in 1905, were "iconoclastic".

Implying that mundane matter – even the ore underground – is stuffed with corked-up energy, this immediately suggested to scientists that our planet could be more like a "storehouse" of dynamite than a sound dwelling.

Early in 1903, Frederick Soddy, one of the founders of nuclear physics, wrote that all that was needed was a tinkering scientist to stumble upon a "suitable detonator" and ignite the "storehouse" by chain reaction. This came in the conclusion of an essay, as he was reaching a crescendo, attempting to awe his readers. (Soddy had a taste for the soaring and dramatic.)

You might also like:

Suggestions planets might explode go back to the Enlightenment, when astronomers started searching for explanations for the abnormally large gap between Mars and Jupiter: a yawning chasm conspicuously filled with debris. Nonetheless, no known force could burst a planetary body. Radioactivity, Soddy conjectured in 1903, might change that.

Not long after Soddy's comments, his research partner Ernest Rutherford made similar noises which were relayed in prominent periodicals. Any "fool", Rutherford had reportedly "playfully" goaded, "in a laboratory", might "blow up the universe unawares". Soddy soon spoke of someone putting their hands on the "lever" that might "destroy the Earth".

This was during the utmost immaturity of nuclear physics. The science was barely over a year old. Such comments, therefore, were baseless conjecture: more rhetorical flare than serious hypothesis. It was almost as if Soddy and Rutherford were bragging, attempting to instil a sense of reverence for, and attract attention to, their nascent field.

Indeed, elsewhere, other researchers were seemingly competing in making increasingly outlandish claims, comparing one granule of uranium, in its latent power, to however many tonnes of TNT. Such flourishes were acts of scientific showmanship: the attempt to awe, and to scare, the public regarding the powers and perils of their Promethean profession. Soddy and Rutherford weren't afraid to use lurid, shocking imagery to do so.

From there, the motif caught hold, taking on a life of its own. Late in 1903, papers were already reporting on the "touch of a key" that could "blow up the whole Earth" through chain reaction. Such claims percolated, and eventually snowballed, through both the scientific community and press. Newspapers spoke of "vast conflagration", of Earth flaring up "in a burst of flame".

In 1909, Irish geologist John Joly added a new ingredient to the myth making. He suggested this might explain the new stars which flare up from time to time in the night sky. Perhaps they are distant planets, others speculated, undergoing atomically induced mishaps. The "outbursts seen in stars", it was supposed, might be the result of "cosmoclastic explosions", brought about by the scientific tinkering of alien civilisations.

As the years went on, serious scientists continued courting the conjecture. Chemist Walter Nernst in 1921 compared humankind to a colony of creatures living upon a gunpowder ball, which would vanish the instant "Prometheus handed them a torch". Charlatans, also, lent the motif momentum. In 1924, one Sheffield University engineer boasted that he was about to successfully crack atoms. (He was good at self-promotion: having previously appeared in papers for inventing a "death ray".) Local journalists sensationalised this, reporting his experiments might "evaporate" not only Sheffield but also the entire Universe. The engineer received terrified letters imploring him not to proceed. "Perhaps if you were a married man with children", one missive read, "you would not be so keen on the possible destruction of the human race".

Such grisly conjectures took on a different tone in ensuing decades, as nuclear weaponry progressively grew into a concrete endeavour and, eventually, a tragic reality. It became deeply distasteful to joke about how Earth's immolation may please "sun-bathers on the beaches of Mars". Yet diverse intellectuals including Carl Jung continued referring to the theory that novae are distant atomic tests gone awry. The journal Science even published a piece on the question, in June 1946, on the eve of the US's post-war tests in Bikini Atoll. It stated that not only "non-scientists" are "disturbed over the prospect" of planetary ignition.

What's to be taken from all this? First and foremost, it shows Teller's fears, alongside those of his Manhattan Project colleagues, didn't emerge in a vacuum spontaneously.

It's perfectly possible they would have arrived at their fear, regardless, had no-one ever imagined similar disasters before. It's also possible their attention was drawn there because the "detonation" motif had seeped so deeply into the cultural consciousness. But this hinged, quite plausibly, on a quip absentmindedly made by Soddy, four decades prior. Back then, Soddy was basing this bombastic quip on not a shred of serious insight; it arose largely from his desire to promote his new field with shock tactics.

He easily could have not included it. Perhaps, if he had not, or if Rutherford hadn't amplified it, it wouldn't have taken root the way it did; perhaps if so, then it wouldn't have accrued the momentum or salience to attract Oppenheimer's attentions.

Our fears are not always the products of pure reason alone. They are also products of certain pasts, where certain motifs get entrenched in the conversation, and others overlooked, in ways hinging upon caprice and chance. A theme can become locked into our consciousness, purely through the self-reinforcement of cumulative familiarity, long before there's any serious research to bear them out.

Present anxieties will be shaped this way too, by motifs we inherit because they flourished previously, often for reasons having little to do with rationality itself. This is cause for thoughtful scrutiny – but not at all for hasty dismissal – when it comes to today's more outlandish-sounding scenarios of AI doom. After all, the case of nuclear fear also shows how a speculative bugaboo can go from science fiction to concrete reality much faster than many expect. In 1929, for example, Nobel-winner Robert Millikan wrote that the energy available through cracking atoms may "perhaps be sufficient to keep the corner peanut or popcorn man going", but "that is all". Many others announced such energy would simply never come.

Nonetheless, we must remain vigilant to the ways cultural tastes can distort our forecasts and fears. Charismatic visions of global conflagration arrived and spread readily for early speculators, but no-one clearly anticipated the more down-to-earth dangers that nuclear explosion heralded in practice. That is, in the form of fallout; the irradiated ash falling, literally, down to Earth, the noxious legacies of which are still suffered by communities – often marginalised ones - today.

Perhaps the lesson for AI is that the dramatic risks should command our attention, but so too should the more tangible, less attention-grabbing, ones. Neither should cancel the other out, especially when – once again – our world is possibly at stake.

--

Join one million Future fans by liking us on Facebook, or follow us on Twitter or Instagram.

If you liked this story, sign up for the weekly bbc.com features newsletter, called "The Essential List" – a handpicked selection of stories from BBC Future, Culture, Worklife, Travel and Reel delivered to your inbox every Friday.