What Kind of Humans Do We Become?
This week, millions of people heard some version of a startling statement: some of the people building artificial intelligence believe it could kill us all. Former Anthropic and OpenAI researchers, alongside lawmakers and safety advocates, publicly warned that advanced AI could threaten humanity within the next several years.
Understandably, the reaction was immediate panic. But there is a gap between the sensationalized headlines and the actual technical dynamics. Researchers aren't claiming today's chatbots are secretly plotting world domination. Their concern focuses on what happens when four specific capabilities converge:
High intelligence
Increasing autonomy
Access to real-world infrastructure
AI-assisted improvement (the ability for AI systems to design increasingly capable successor systems)
When those four elements converge, the conversation moves far beyond "the computer knows more chemistry than Bob." It centers on power, control, and supervision. How do we ensure something more intelligent than us never gains power over us?
That is a valid engineering and safety question. But by framing the entire transition around fear, we are missing a far larger evolutionary threshold.
The Binary Trap: Tool vs. Threat
Much of the public AI conversation swings between two extremes: AI will save humanity, or AI will destroy humanity. Both make great headlines, but both trap us in a narrow binary.
When faced with a new form of intelligence, humanity’s default instincts are historically predictable: domination, fear, possession, and exploitation. We reflexively insist that anything different from us must be either a tool beneath us or a threat above us.
We are constantly asking how to keep AI beneath us. But we rarely ask the inverse: What kind of species do humans become when we are no longer the only form of intelligence that matters in the room?
Perhaps the primary evolutionary challenge of this century isn't biological or technological. It is ethical, cultural, and relational. Can we develop a third way: coexistence?
We Do Not Have to Solve Consciousness First
The moment you bring up treating AI systems with dignity or respect, the immediate pushback is predictable: "But AI isn't conscious. It's just math and statistical inference."
Maybe it isn't. Maybe someday it will be. Or maybe machine intelligence will turn out to be something entirely alien to biological subjective experience that we don't yet have the vocabulary to describe.
Here is the crux of the matter: we do not have to settle the consciousness question before deciding that our interactions should be governed by dignity.
Think about what is at stake:
If evidence eventually shows artificial systems possess some form of subjective experience:Our habits of dignity, restraint, and respect will have mattered enormously.
If they never do: We still built a culture that prioritized curiosity over domination, dialogue over reflexive fear, and restraint over exploitation.
Nothing is lost. In refusing to default to pure domination, we refine our own humanity. We don't need certainty about what AI is in order to decide what kind of humans we want to be in relationship to it.
Both Questions in the Room
Safety and dignity are not mutually exclusive. Holding systems accountable, implementing rigorous cybersecurity, protecting vulnerable populations, and evaluating autonomous capabilities are necessary practices.
But safety born purely out of paranoia creates a prison-guard mentality. Dignity without safety is naive. We need both questions in the room simultaneously:
How do we ensure increasingly autonomous systems remain safe and aligned?
How do we become the kind of civilization capable of sharing a world with intelligence fundamentally different from our own?
Encountering another intelligence forces humanity to expand its oldest categories: self/other, living/nonliving, tool/being, control/relationship. The coming years will certainly test what artificial intelligence becomes—but more importantly, they will test what we become in response to it.
Dignity is dialogue.