For billions of years life on Earth has followed a simple rule: survive, adapt, compete, or disappear.
Dinosaurs once dominated the planet. They are gone. Countless species rose, evolved, competed and vanished and eventually humans emerged with something that changed the balance completely: intelligence.
Now, for the first time we are deliberately trying to create intelligence outside ourselves. And that raises a question that sounds like science fiction but is becoming increasingly difficult to dismiss: what if the next dominant intelligence on Earth is something we are building ourselves?
That thought has become harder to ignore after a series of extraordinary warnings from inside the AI industry. Earlier this month, former Anthropic researcher Jacob Coxon resigned and publicly warned that people building advanced AI genuinely believe it could potentially kill humanity by the end of this decade. Evan Hubinger, Anthropic’s Alignment Science Lead, went even further, saying he personally puts the probability of AI killing all humans within the next decade at more than 10 percent. These are personal assessments, not scientific predictions or established facts. But the fact that such warnings are coming from people who have worked directly on frontier AI systems has intensified an already growing debate about where this technology is heading.
And the warnings haven’t stopped there. Former Google DeepMind researcher Bilal Chughtai has also gone public with concerns about the possibility of AI posing an existential threat to humanity, arguing that the technology may be advancing faster than our ability to make it reliably safe.
For a while, this debate seemed to exist mainly between AI optimists and AI safety researchers. Then something unusual happened. Anthropic CEO Dario Amodei called for the industry to pace the development of frontier AI. His argument was not that AI should be abandoned but that increasingly powerful systems need stronger safety mechanisms, independent evaluation, and greater caution before they are deployed. OpenAI CEO Sam Altman and Elon Musk also publicly expressed support for the broader idea of slowing the pace.
Think about how unusual that is. Some of the people building the most powerful AI systems in the world are now saying, in effect that perhaps we need to slow down.
Because AI is no longer simply answering questions or generating images. It can code, use tools, conduct research, operate as an agent, and increasingly perform tasks with less direct human involvement. The concern among AI safety researchers is not that today’s chatbot will suddenly decide to destroy humanity. It is about what could happen if future systems become substantially more capable, autonomous, and difficult to control. And that is where the question of evolution becomes interesting.
Human intelligence gave us the ability to reshape the planet. We developed agriculture, electricity, nuclear weapons, biotechnology, and the internet. We became the dominant species not because we were physically the strongest, but because intelligence gave us an extraordinary ability to solve problems and build tools. AI could represent another step in that process. We are building systems designed to perform tasks that once required uniquely human abilities like reasoning, programming, research, creativity, and decision-making.
So what happens if the thing we create eventually becomes much better at those things than we are? The answer to this is “We don’t know.” And that uncertainty may be the most important part of this entire debate. There is no scientific consensus that AI will inevitably destroy humanity. The extinction probabilities being discussed publicly are individual judgments and experts disagree dramatically about how likely such outcomes are.
The United States and China are competing for leadership in artificial intelligence. Companies are competing for talent, computing power, investment, and market dominance. Governments want the economic and strategic advantages that advanced AI could bring. President Donald Trump, meanwhile has dismissed fears of AI catastrophe as a “hoax” and has repeatedly framed AI leadership as a strategic competition with China. That reflects a very different calculation: slowing down could mean surrendering technological advantage to a rival. And perhaps that is the real paradox. Everyone wants AI to be safe. But nobody wants to be the first to slow down.
We have seen a similar dynamic before. Nuclear weapons gave humanity an extraordinary new power and immediately turned that power into a competition between nations. AI is obviously not the same technology, but the geopolitical logic is familiar that whoever gets there first may gain an enormous advantage. That creates an uncomfortable possibility.
The real challenge of the AI age may not be whether we can create machines that are smarter than us. The harder question is whether our wisdom, institutions, and ability to cooperate can develop fast enough to keep up with the intelligence we are creating.
Because maybe the most extraordinary achievement of human intelligence will not be creating something smarter than ourselves.It will be knowing what to do once we have.