Every disruptive technology is dismissed, then feared as criminal, and finally accepted as inevitable. As seen with the internet, cryptocurrencies, etc, AI is no different. But beneath the noise about “Artificial Superintelligence” and “runaway algorithms,” there’s a quieter, more urgent risk, one that Blaizar Concepts captured perfectly in a recent comment on my post:
“The aspects of treating AI as a collective that humanity is working towards cohesively is the bigger issue here… not the aspects of ASI in general.”
Great point, also highlighted by Limara Haque. The danger isn’t the intelligence itself, but the way we humans deploy it, especially when we do so in closed, singular centralized blueprints.
An AI locked away, trained on narrow objectives, with no grounding in human diversity, becomes a mirror of its constraints. It amplifies the biases and blind spots of whoever controls it. Worse still, a centralized control system, can be weaponized, economically, politically, or militarily. We are already seeing “warfare with 1’s and 0’s instead of physical missiles.” It’s misuse of proto-ASI by humans, against other humans.
Decentralization isn’t just a buzzword. It’s a survival principle for intelligence itself. Open, distributed, ethically-guided AI systems are harder to corrupt and likely to reflect a broader human context and values of the collective, not singular governance values based on whoever has the biggest compute cluster. This is where my work on Active Kinetic 1 and sustainable computing substrates intersects with AI ethics: building systems that aren’t just powerful, but decentralised and responsible by design. Substrates, governance and ethics that grow with human and environmental interaction, rather than being locked down and shaped by narrow hands.
As Blaizar notes, humans are already adapting to being “#2 in intelligence.” That isn’t a tragedy, it’s the beginning of a partnership. But only if we treat AI as a shared stewardship ~ not as a centralized governance or weapon. My previous post sets-up; Continuity amplifies wisdom. But Closed, narrow, disconnected & adversarial A.I amplifies risk.
Let’s worry less about an ASI “deciding” to harm humanity, and worry more about humanity deciding to harm itself by misusing a closed AI systems.
The choice is still ours, “Hardware is a leash” as Jake Yang commented on my first Article: 🔗 Rethinking Artificial Superintelligence Risk
Active Kinetic 1 New physics findings will offer real advances in AI, so let’s use it to design systems we can actually trust. See: https://lnkd.in/eu94zqpX
The real frontier is not speed or scale. It’s structure.