ArtVersion Principal and Creative Director Goran Paun is featured in the Senior Executive Think Tank panel “Building a More Competitive and Safer AI Ecosystem,” published June 18, 2026. The panel asks AI and digital leaders a single question: what one rule they would change to make AI competition more open, trustworthy, and safe.
Where much of the discussion centers on data portability, open training corpora, and infrastructure access, Paun takes a different angle. His concern is that the current race over-rewards speed, namely who can generate, automate, and ship the fastest. “I would build in a generative pause,” Paun says. The idea is not a pause on innovation but a practical checkpoint before AI outputs move straight into business, legal, creative, or operational decisions, with clearer provenance and accountability around the work. The advantage, in his view, belongs to the organizations that can tell which outputs to rely on, which to challenge, and where a human call still belongs.
The Generative Pause
This was not an offhand position. The generative pause is the subject of a VERSIONS® white paper, “The Generative Pause: A Framework for Designing Human Judgment Into Digital Systems,” first published in early 2025 and co-authored by Paun and Erin Lentz, Executive Director of Design at ArtVersion. The paper makes the case at length: three decades of design optimized for speed, the frictionless flow that carries a user from intention to action, were built for low-stakes choices and do not hold up once interfaces begin mediating financial commitments, clinical outputs, legal authorizations, and personnel decisions.
The framework names that gap the velocity problem and answers it with three governing principles. Conscious Verification replaces passive approval with structured engagement, so a person has to show real contact with an output before the system proceeds. Symmetric Benefit holds that a pause must give the user what they need to decide well, beyond giving the organization a record that approval occurred. Proportional Calibration matches the depth of the pause to the actual weight of the decision, since a routine sign-off and an irreversible authorization should not carry the same friction.
The argument is grounded in behavioral economics, in dual-process theory and hyperbolic discounting, and in where regulation is heading, including the EU AI Act’s treatment of human oversight as a design requirement rather than a matter of record. It also reframes how to measure success, proposing comprehension rate, error interception, downstream remediation cost, and long-term retention as the signals that show whether a pause is doing real work, rather than completion speed on its own.
Read the full discussion on Senior Executive.