Artificial intelligence has crossed a threshold. It is no longer an experiment, nor merely a tool. As Dario Amodei argues in The Adolescence of Technology, AI now resembles something more unsettling and familiar: a powerful adolescent that is capable, unpredictable, idealistic, and insufficiently governed.
This framing matters. Adolescence is not simply about growth; it is about mismatch – between capability and wisdom, power and restraint, intention and consequence. AI today sits squarely in that gap.
We are no longer asking whether AI will shape civilization. We are deciding how it will mature – and whether we will mature alongside it.
From Childhood to Adolescence: The End of Innocence
In earlier stages, AI could be treated as a novelty or productivity enhancer. That era is over.
As explored in The Rise of the AI Overlords: Humanity’s Childhood’s End Moment, humanity is exiting a period of technological innocence. AI systems now reason, plan, persuade, generate strategy, and increasingly act autonomously within economic, informational, and political systems.
This is not science fiction. It is structural reality.
Amodei’s essay sharpens this insight by emphasizing that capability is accelerating faster than governance, norms, and institutional adaptation.
Adolescents can drive cars and wield influence before they fully grasp consequences. AI can now “drive” major systems – markets, narratives, logistics – while our guardrails lag behind.
The Risk Is Not Malice – It Is Misalignment
One of the most important clarifications Amodei makes is that the danger of AI does not stem primarily from hostility. It stems from scale without alignment.
AI systems do not need intent to cause harm. They only need:
- poorly defined objectives
- asymmetric incentives
- unchecked deployment
- or integration into fragile human systems
This mirrors themes explored in The Great Human–AI Convergence, where the central question is not whether humans and AI will integrate – but whether that integration elevates human agency or dissolves it.
Adolescents often believe they understand the world better than they do. So do institutions deploying AI at scale without fully understanding second- and third-order effects.
Power Without Governance Is Not Progress
A central warning in The Adolescence of Technology is that technological maturity does not automatically produce social maturity.
We have seen this before:
- Industrialization outpaced labor protections
- Financial engineering outpaced systemic risk controls
- Social media outpaced democratic resilience
AI now risks repeating the pattern at far greater speed and magnitude.
This is why governance cannot be an afterthought.
As argued in Governance at the Edge of Intelligence, the coming era will demand new leadership models – ones that understand intelligence itself as a strategic resource requiring stewardship, not exploitation.
Governance here does not mean centralized control. It means:
- transparency over opacity
- accountability over abstraction
- human-aligned incentives over pure efficiency
Without this, adolescence curdles into recklessness.
Where Are We Now in the Human–AI Future?
Given what we know today, we are likely in the early-to-mid adolescence phase of the human–AI relationship.
Characteristics of this phase include:
- Rapid capability growth
- Uneven institutional response
- Cultural fascination mixed with anxiety
- Economic disruption without clear redistribution models
- Competing visions of control vs. partnership
We are not yet at irreversible outcomes, but the window for shaping trajectory is narrowing.
If we succeed, the next phase resembles mature partnership:
- AI as amplification of human judgment, not replacement
- Distributed governance, not technocratic dominance
- Economic systems redesigned for abundance rather than scarcity
If we fail, the alternative is not dystopia overnight – but gradual erosion:
- concentration of decision-making power
- loss of meaningful human agency
- systems optimized for metrics, not values
Adolescence does not guarantee adulthood. It only offers the possibility.
Choosing the Direction of Maturity
Amodei’s essay is ultimately a call for intentional adulthood – for slowing down where necessary, thinking systemically, and refusing the illusion that capability alone equals progress.
The question before us is not whether AI will grow up.
It will.
The real question is whether we will grow with it, or allow ourselves to be governed by systems we no longer meaningfully understand.
History suggests that civilizations are remembered not for the power of their tools, but for how wisely they learned to wield them.
The adolescence of AI is ending.
What comes next is still open to human direction and control… but not for much longer.