A new model is released. It is more powerful than the previous one. The regulators are alarmed. They draft new rules. They debate the rules. They pass the rules. The rules are outdated. The model has already evolved. The regulators are behind. They are always behind. They cannot keep up. This is the regulation that will fail.
The EU AI Act is a landmark. The US executive orders are ambitious. The global efforts are well-intentioned. They are all doomed.
The Pace of Progress
The pace of progress is too fast.
The Concept:
AI is evolving rapidly.
Regulations are slow.
The gap is growing.
The Consequence:
Regulations are outdated.
They are not effective.
They are not enforceable.
A Contrarian Take: The Pace Is Not the Problem. The Approach Is.
The pace is not the problem. The approach is. The regulators are using the wrong approach.
They are trying to regulate the technology. They should regulate the applications.
The EU AI Act
The EU AI Act is a comprehensive framework.
The Concept:
It classifies AI systems by risk.
It imposes requirements on high-risk systems.
It bans unacceptable practices.
The Problem:
It is rigid.
It is slow.
It is not adaptable.
A Contrarian Take: The EU AI Act Is Not a Failure. It Is a Starting Point.
The EU AI Act is not a failure. It is a starting point. It is the first of its kind.
It will evolve. It will improve.
The US Executive Orders
The US executive orders are ambitious.
The Concept:
They direct agencies to develop guidelines.
They promote safety and security.
They encourage innovation.
The Problem:
They are not binding.
They are not enforceable.
They are not comprehensive.
A Contrarian Take: The Executive Orders Are Not a Failure. They Are a Signal.
The executive orders are not a failure. They are a signal. They show the direction.
They will be followed by legislation.
The Global Efforts
The global efforts are fragmented.
The Concept:
Different countries have different approaches.
They are not coordinated.
They are not consistent.
The Problem:
There is no global framework.
There is no global enforcement.
There is no global coordination.
A Contrarian Take: The Global Efforts Are Not a Failure. They Are a Beginning.
The global efforts are not a failure. They are a beginning. They are the first step.
They will lead to global coordination.
What Might Actually Work
Regulation can work if it is designed differently.
- Adaptive Regulation:
The regulation should adapt to the technology.
It should be flexible.
It should be iterative.
- Application-Based Regulation:
The regulation should focus on applications.
It should not focus on the technology.
It should focus on the use.
- Multi-Stakeholder Governance:
The regulation should involve all stakeholders.
It should involve developers, users, and civil society.
It should be collaborative.
A Contrarian Take: Regulation Will Never Work. Self-Regulation Will.
Regulation will never work. Self-regulation will. The industry will regulate itself.
The market will force responsibility.
What This Means for You
You are a user of AI. You have a stake in regulation.
- Be Informed:
Be informed about the regulations.
Be informed about the debates.
- Be Engaged:
Engage with the process.
Provide feedback.
- Be Responsible:
Use the technology responsibly.
Be part of the solution.
The Last Regulation
The last regulation is not a law. It is a choice.
You ask: "Can AI be regulated?"
The AI says: "It depends."
You realize: The regulation is not in the law. It is in the use.
If you could design a regulation that works, how would you do it? And why?
Top comments (0)