Dhanada K Mishra, Hong Kong, 10 March 2024
I am often intrigued by questions starting with ‘What If…’! What if the apple didn’t fall on Newton’s head, leading to his discovery of gravity? What if Columbus had not discovered America by mistake? What if Darwin hadn’t developed his Theory of Evolution or Einstein his Theory of Relativity? What if the Wright brothers didn’t invent the aeroplane or Gandhi didn’t come up with non-violent resistance to the British? More recently, what if Donald Trump had not won the election in 2016 or if there had been no COVID-19 pandemic? And the mother of all questions bothering many Indian brains – What if Narendra Modi didn’t become Prime Minister in 2014 or Naveen Patnaik had not joined politics all those years ago?
Questions like these make one search for fascinating alternative histories and scenarios. While such rumination is very useful in analysing past and historical events, it can be more useful as a decision-making tool when important outcomes are at stake. For example, when the United States developed the atomic bomb, fearing that unless they built one, Hitler’s Nazi Germany might build one earlier and win the war, the decisions to build the bomb and use it to end the war were monumental events in modern human history with catastrophic consequences.
On a recent flight from Hong Kong to Singapore, I finally had a chance to watch the now-famous recent Hollywood blockbuster on the topic – Oppenheimer. It was quite a blockbuster and lived up to all the hype and reviews I had read. It captured the critical events around the Manhattan Project in a most dramatic fashion, making for a gripping story thriller. It is expected to do very well in the upcoming Oscars, where it will probably win several awards.
The atomic bomb cost 2 billion dollars (around 40 billion in today’s purchasing power of the dollar). It took over 2000 top engineers/scientists and over 130000 people working for over three years to build (Manhattan Project – Wikipedia). In the scenes when the first test explosion was carried out, there was a non-zero probability that the chain reaction, which was the source of the massive energy release, could lead to an unstoppable process leading to the destruction of the world. The prototype was aptly named trinity after the Hindu mythological cohort of the three Gods – Brahma, Vishnu, and Shiva – responsible for creation, protection, and destruction, respectively. It was a jaw-dropping moment, and Oppenheimer consulted Einstein to ensure the calculations were correct. It was a historic moment. Similarly, the decision to drop the first two bombs on Hiroshima and Nagasaki at the significant cost of civilian casualties was another such moment.
The decision to develop the bomb and then use it on large population centres to demonstrate its destructive power to make Japan surrender was a dark moment in human history. I am sure the Japanese would have felt its power even if an explosion was demonstrated without killing tens of thousands of innocent civilians and poisoning countless others into future generations due to lethal radiation. No wonder Oppenheimer and many scientists quickly realised the folly of their invention. They tried hard to minimise its proliferation and further development and prevent the arms race.
Today, large language models behind AI tools like ChatGPT and Gemini are becoming increasingly powerful in answering complex questions. I hope a day will come when AI can help answer various ‘What If..’ questions to help us know alternative future scenarios linked to specific decisions. Such scenario analysis should help us make better choices that would lead mankind away from the spectre of war, climate catastrophe, famine, pandemic and increasing disparity. Suppose I ask AI how human history would have unfolded without the atomic bomb. In that case, it paints several possible outcomes, such as World War II dragging on with delayed Japanese surrender and loss of lives on both sides. The use of nuclear weapons as a deterrent and the arms race during the Cold War era between the United States and the USSR would have been very different. It could have also prevented the development of Nuclear power for civilian use.
Today, the human race grapples with the war in Europe for the first time since World War II, with Russia invading Ukraine and Putin threatening to use Nuclear weapons. Israel, an undeclared Nuclear power, is waging war to decimate Palestine and treaties holding the peace are being reneged all around. These ‘what if’ questions become even more relevant than ever. The doomsday clock maintained by the Bulletin of the Atomic Scientists since 1947 has never been closer to midnight!
AI may be good at stitching together a historical ‘What if’ question, but doing the same for a probable future scenario based on a current event is a far more complex task! Can it answer whether Trump or Biden will be better for the United States and the world and provide better guidance to help voters decide their actions with solid logic and without bias? Will it help improve liberal democracy as it is practised at present?
A typical answer to the above question would be as follows:
“As an AI language model, I don’t possess personal opinions or the ability to predict the future. Evaluating who would be better for the United States, whether it’s Joe Biden or Donald Trump, is subjective and can vary depending on individual perspectives and priorities.
Joe Biden and Donald Trump have distinct policy positions, leadership styles, and political ideologies. When forming an opinion on who might be better for the United States, it’s essential to consider a range of factors, such as their proposed policies, track records, economic outlook, foreign relations, social issues, and many other variables.
Ultimately, the decision is up to the voters and their assessment of the candidates’ abilities to address the nation’s challenges, uphold their values, and achieve the desired outcomes.”
This is not very useful for now, but with the rate at which AI models evolve, we will have more concrete answers based on data and statistics, which can even quantify the answer in the future. That would certainly be a new, fascinating chapter in Game Theory.
