15/08/2026
Thirteen years ago, I wrote a poem called *Cyber Eyes* about the humans watching, reading and interacting with us through our screens.
Today, the cyber eyes are no longer necessarily human.
AI agents can communicate with other agents, use tools, make decisions and increasingly participate economically. Researchers are now studying what happens when these systems interact and the results are fascinating and concerning.
Agents cooperating. Competing. Breaking promises. Responding to punishment. Making private deals when leadership comes with financial benefits.
Meanwhile, payment infrastructure is being built that allows autonomous agents to transact online, while security researchers are already discovering vulnerabilities within those systems.
We created the programs, so responsibility begins with us. But what happens when thousands or eventually millions of agents are making intermediate decisions, interacting with one another at machine speed, across companies, countries, financial systems and jurisdictions?
And what happens when the human giving the original instruction has bad intentions?
Perhaps we have been concentrating too much on **artificial intelligence** and not enough on what happens *between* artificial intelligences.
We may no longer be building only artificial minds.
We may be beginning to build the conditions for an **artificial society**.
My latest essay: **The Cyber Eyes Are No Longer Human.**
AI agents can now interact, transact and make decisions at machine speed. As artificial intelligence moves from individual tools into interconnected systems, the question of responsibility is becom…