Get Muhammad Arnob on Ai Ai Ai

As AI systems become more capable, decisions about deployment, safety, and control become increasingly consequential. Should that responsibility rest with companies, independent researchers, or global institutions? What checks and balances would you advocate for?

There is a clear trend toward large language models, but there are also quieter movements — mechanistic interpretability, small models, retrieval-based architectures. Which of these should we be paying more attention to, and why?