Expert Takes: Dan Nechita

Dan, you have been directly involved in the genesis and negotiations of the EU’s AI Act, the world’s first legislation on Artificial Intelligence. Can you explain how this pioneering regulation will now be implemented?

The AI Act represents a new type of regulation, not only because of its development process but also because it is grounded in the core values of the Union—democracy, individual freedoms, fundamental rights, and more.

Regarding its implementation, the Commission will need to issue guidelines in the coming months to make the legal requirements of the Act more practical. Another crucial aspect involves the work assigned to standard-setting bodies, which will be essential for understanding how to implement the Act at a technical level. However, what’s really important—especially when considering the Metaverse and the convergence of various technologies—is the establishment of the AI Office. This office will serve as the central European enforcer for very powerful general-purpose AI models, particularly those with systemic risks, which might be the type of AI deployed in a Metaverse environment.

Looking back, it’s interesting to note that the AI Act was initially intended as a product safety regulation, but we ended up focusing on high-risk use cases. This is a powerful approach because it genuinely considers how technology is used and its impact on end users or those interacting with it. We’ve established a flexible governance structure to monitor the frontier of technology, featuring an interactive component with the AI Office supervising the market. Essentially, it acts as a European supervisory authority, enabling flexible enforcement through model analysis and industry empowerment. This gives me confidence that the regulation will be future-proof. We’re not over-regulating; instead, there’s a dynamic interaction between industry and the regulator.

Overall, I believe that policymakers today are responsible for less and less, and it’s the people building our digital future who need to carefully consider their creations. The technology they develop will be with us for a very long time; once it’s here, it won’t disappear. This enduring presence brings with it a profound responsibility. While the potential is enormous, so too is the moral responsibility, for both AI and Metaverse developers.

What specific opportunities and challenges do you see being associated with the Metaverse. And are these going to call for regulation too?

While I believe in the potential of the Metaverse, it’s important to remember that this isn’t the Matrix; it hasn’t reached societal level integration yet. We shouldn’t scare ourselves unnecessarily or rush into regulation. Currently, regulatory needs address the entire digital environment, tackling issues like cybersecurity, identity theft, manipulation, and targeted advertising. These concerns are being managed. However, the tools and capabilities being developed now will shape our digital future. Soon, technologies like the Metaverse, quantum computing, and AI will converge, creating a new paradigm.

Regarding the Metaverse’s opportunities and challenges, you could imagine this is a sandbox for exploring and testing new technologies or systems. For instance, using AI to distribute tokens in a virtual environment might be safer than allocating state allowances to underprivileged families. The Metaverse can recreate the real world with different rules, offering a space for human creativity, beyond the limitations of the real world.

Social interaction in the Metaverse could also evolve positively. If we can build virtual worlds with better business models than ad-driven social media, we could create environments that foster genuine social connections and community building. This could be a remedy for the shortcomings of current social media platforms, encouraging more meaningful interactions.

Finally, the Metaverse presents complex questions about democracy and individual freedoms. In a virtual world, voting might be blockchain-based, but who sets the rules for voting? Who runs the system? Ensuring that the Metaverse facilitates democracy requires vigilance. It could potentially enable a greater extent of democracy than we have today, but this outcome is not automatic. Democracy is a complex system built over centuries, and the same care must be taken to establish its principles in the Metaverse.

How will the EU’s new AI Act apply to the Metaverse?

Considering the end goal of a large-scale Metaverse, it makes sense that the AI behind it would be a very powerful general-purpose model, akin to those currently being trained at the cutting edge of technology. While we hope these models will not be risky or dangerous, they will likely fall under the purview of ensuring their safety, a task designated to the AI Office.

I also believe that in the not-so-distant future, all these technologies—Metaverse, AI, quantum computing— will merge and intersect. Currently, there’s a lot of experimentation at these intersections, but eventually, they will work seamlessly together. This integration will unlock new possibilities and foster greater creativity.

As I was saying earlier, with this immense potential comes great responsibility for metaverse developers.

This interview is from our first opus, released in June 2024. You can read the full opus here on our website.