His reasoning is based on an annual probability. He cites an estimate of around 1 per cent per year during the Cold War, and then puts the current risk at around 2 per cent per year – that is, a one-in-fifty chance each year. He himself describes this calculation as ‘rough’ and acknowledges that it is not a rigorous assessment. By applying this risk year after year, he arrives at an average timeframe of around 35 years before a nuclear war. This average does not mean that a catastrophe will necessarily occur by that point: it reflects the way in which a cumulative risk can build up over time.
Why this concern? Gross cites the weakening of arms control agreements, the proliferation of nuclear powers and international tensions. He also fears the increasing automation of weapons systems. If critical military decisions were to be taken by software or artificial intelligence systems, he argues that an error could accelerate an escalation before a human could intervene.
The phrase has caught on easily because it condenses a long-standing anxiety into a dramatic deadline. However, the value of the interview lies less in the idea that ‘the end is nigh’ than in the political warning it conveys: nuclear arsenals still exist, and the danger they pose also depends on human decisions. Gross also emphasises that this risk can be reduced through dialogue and concrete measures between states.
This statement must therefore be read for what it is: a wake-up call from a concerned physicist, based on a debatable estimate rather than on a consensus predicting the extinction of humanity in fifty years’ time. The statement is frightening; the question it raises is more useful: how can we prevent a risk that is low on an annual basis from becoming, through accumulation and recklessness, a major threat to future generations?
