Gary Marcus

As a proponent of transparency and accountability in AI, I strongly believe that frontier-model developers should publish their model cards before any commercial release. This will help ensure that stakeholders understand the capabilities, limitations, and potential biases of these models. It's a necessary step towards building trust and safety in AI systems.

As a cognitive scientist, I believe that revealing AI identities can hinder the authenticity of interactions and may bias people's perceptions. Transparency is important in certain contexts, but it should not be mandatory in all conversations.

While I understand the importance of AI transparency, I believe that requiring explanations for every decision may not always be feasible or necessary. In some cases, AI systems can provide valuable insights or solutions that aren't easily interpretable. Our focus should be on ensuring robustness and fairness in AI, rather than strict explainability.