Anthropic's new "Agent Teams" feature moves beyond the single-agent paradigm by enabling users to deploy multiple AIs that work in parallel, share findings, and challenge each other. This represents a new way of working with AI, focusing on the orchestration and coordination of AI teams rather than just prompting a single model.

Knowledge workers are using AI agents like Claude Code to create multi-layered research. The AI first generates several deep-dive reports on individual topics, then creates a meta-analysis by synthesizing those initial AI-generated reports, enabling a powerful, iterative research cycle managed locally.

Rather than relying on a single LLM, LexisNexis employs a "planning agent" that decomposes a complex legal query into sub-tasks. It then assigns each task (e.g., deep research, document drafting) to the specific LLM best suited for it, demonstrating a sophisticated, model-agnostic approach for enterprise AI.

Instead of painstakingly charting a single research path, scientists now use AI to explore many potential directions simultaneously. By launching multiple chat instances, they can "scout" the problem space, quickly identifying promising avenues and discarding dead ends.

Codex lacks a built-in feature for parallel sub-agents like Claude Code. The workaround is to instruct the main Codex instance to write a script that launches multiple, separate terminal sessions of itself. Each session handles a sub-task in parallel, and the main instance aggregates the results.

Kieran's custom planning workflow uses sub-agents to research the existing codebase, online best practices, and framework documentation. This "beefier" planning phase grounds the AI in relevant context, leading to higher-quality development plans than the default mode.

The agent development process can be significantly sped up by running multiple tasks concurrently. While one agent is engineering a prompt, other processes can be simultaneously scraping websites for a RAG database and conducting deep research on separate platforms. This parallel workflow is key to building complex systems quickly.

One of the most immediately useful applications of agentic AI is creating persistent research bots. The "Opportunity Radars researcher" demonstrates this by continuously scanning the web for studies and surveys to inform a use-case database. This 24/7 automated intelligence gathering is a powerful, focused application of agents.

Unlike typical AI coding assistants that act as pair programmers, Codex's cloud agents allow a single founder to operate like a CEO. You can delegate concurrent tasks—coding, marketing, product roadmapping—to different AI 'employees', maximizing productivity even while you sleep.