Responsible Progress: A story of hope

There is a photograph of J. Robert Oppenheimer taken shortly after the bombing of Hiroshima. He does not look like a man who has triumphed. He looks like a man who has understood something terrible about himself and cannot forget it. Years later, he would articulate from Hindu scripture what that photograph plainly displayed: “Now I am become Death, the destroyer of worlds.” He had built the most destructive weapon in human history, and he spent the rest of his life carrying that weight. 

We are, right now, in the years before that photograph is taken again. 

The pursuit of a machine that can think, reason, and act across every domain of human knowledge – is being run at a pace that would make the Manhattan Project look cautious. The people running hardest are not, for the most part, philosophers or ethicists or community leaders. They are technologists and investors, many of them brilliant, some of them genuinely well-intentioned, almost all of them moving too fast to ask the question that matters most: what happens to people? 

Oppenheimer is far from being history’s only cautionary inventor. Alfred Nobel, who invented dynamite intending it to revolutionise construction and mining, watched it become the instrument of industrialised slaughter. He was so disturbed by his creation that he put forward the Peace Prize in his name. An act of conscience by a man trying to rewrite what his name would be remembered for.  

Mikhail Kalashnikov, designer of the AK-47, reportedly spent his final years tormented by the deaths attributed to his rifle, writing to the Russian Orthodox Church asking whether he bore moral responsibility for those killed by it. Even the Wright brothers, who gave humanity flight, lived to see their invention transformed into a platform for bombing cities. 

The pattern is consistent and heartbreaking. A brilliant mind solves a technical problem. The solution spreads. Those with the most to gain from weaponising it – financially, militarily, politically – are sufficiently motivated and funded to get there before those who might wield it wisely. The inventor is left watching from the sidelines, their creation doing things they never sanctioned or wanted. 

The honest answer is that they are largely optimising for capability. For benchmarks. For the moment their model beats the previous record on some abstract measure of intelligence. The question of whether that capability serves human flourishing, whether it strengthens communities, whether it preserves the dignity and agency of ordinary people. That question is, at best, secondary. At worst, it is seen as an obstacle. 

This is not cynicism. It is pattern recognition – something which, ironically, AI is incredibly good at. 

We have already seen what happens when technology scales faster than wisdom. Social media promised connection and delivered polarisation, addiction, and the systematic dismantling of shared reality. Search engines promised access to knowledge and delivered misinformation ecosystems optimised for engagement over truth. The engineers who built these systems were not evil. Most of them believed they were making the world better. But the profit motive, the competitive pressure, the simple intoxication of whether we can, consistently drowned out the harder, slower question of whether we should. 

The recent publication by OpenAI of their “Industrial Policy for the Intelligence Age: Ideas to Keep People First” appears to have the right intent, and by its own admission are “starters for discussion”. Sadly, rather than enter into those discussions, recent news showed that one individual had turned towards extreme violence against Sam Altman. We want make it very clear here, and without ambiguity, that violence is not the right path to take. Discussion with a common goal is. But the longer that people remain threatened by the technology, and the more jobs that are openly lost because of AI adoption, the more that negative sentiment will grow. The OpenAI document is a wish list – some things easy, some things harder, some naïve – but a good place to start.  

The stakes are fundamentally different. A system capable of outthinking humans in every domain: medicine, law, science, governance, warfare, in the hands of those who prioritise power or profit over people, is not a productivity tool. It is a civilisational risk. 

This is why documents like the Human-First AI Adoption Manifesto matter so urgently, and why they need to be more than aspirational wall art. 

The manifesto is built on a core principle so obvious it should not need stating: Humans decide, AI assists. From the intent of any AI adoption, through the design of the solution, the operation of the system created, and the ongoing enhancement, it embeds human oversight and human priorities above everything else. AI should augment human capability, never erode human judgment or agency. 

These are not radical ideas. They are the ideas of people who have looked at the history of transformative technology and drawn the sensible conclusions. 

What makes the current moment both dangerous and hopeful is that we are not yet in the aftermath. We are still in the years when choices matter, when the trajectory can be bent, when the people building these systems can be held (and can hold themselves) to a higher societal standard.

The investment cycles are accelerating. The competitive dynamics between nations and corporations create pressure to move faster, to deploy wider, to ask fewer questions. Every week of delay feels, to those in the race, like a week of advantage surrendered to a rival. This is precisely the logic that built the atom bomb, and built it again, and again, until there were enough warheads to end human civilisation several times over. 

We need to break that logic. Not by stopping progress, though. AI genuinely can cure diseases, extend human potential, connect isolated communities, and solve problems that have defeated us for generations. The technology is not the villain. The absence of wisdom in its deployment is. 

The inventors who lived with regret did not regret inventing. They regretted the world they handed their inventions to, the insufficient thought they gave to who would benefit and who would suffer, and the governance mandated to ensure a wise use of their gift. 

We do not have to repeat that. We still have the chance to influence the direction of this technology through our choices of what we adopt and what we avoid. Choose to bolster rather than banish your people.  

The manifesto already exists. The principles are written. The only question left is whether we have the collective will to make them mean something before it is too late to matter. 

Darwin Lee,
April 2026