In this commentary, the speaker argues that the primary question when selecting a video API today should be whether the vendor is prepared to support AI agents, as the majority of future video minutes will involve human‑to‑AI interactions. He outlines four video meeting archetypes—1‑to‑1, group, broadcast, and 1‑to‑LLM (human speaking to a large language model)—and emphasizes that the 1‑to‑LLM archetype will dominate future usage. The speaker stresses that agentic AI is not a separate add‑on but an integrated layer that will appear across all archetypes, initially manifesting as AI scribes, note‑takers, co‑hosts, and live translation services. Consequently, the key decision is not which archetype a company needs, but whether a video API vendor can deliver "N‑Jetting" AI capabilities to the chosen archetype at the right time. The speaker’s upcoming report will evaluate which vendors are building credible N‑Jetting AI and which will force customers to retrofit later.
