In the race to make artificial intelligence faster, a new perspective argues that slowing down could actually be the key to building trust. Raji, a commentator on the YouTube channel Trust Issues Show, suggests that friction, latency, and deliberate pacing in AI interactions—especially with agentic AI—are exactly what make the technology safer and more trustworthy.
"With agentic AI, that loading screen might be the most important feature of all," Raji explains. The idea is that by building in deliberate pauses, users have time to review and approve AI actions, reducing the risk of unintended consequences. Instead of viewing delays as flaws, this viewpoint repositions them as critical safety features.
The full episode explores how slowing down AI can enhance transparency and user control, ultimately winning over skeptical users. As AI systems become more autonomous, the argument goes, haste could lead to mistakes; patience could build reliability.