To maintain the high velocity enabled by AI agents, you must be available to unblock them constantly. This means making critical decisions and even landing pull requests from your phone is no longer a novelty but a competitive necessity. Delaying until you're at a desk is a major disadvantage.
Frontier model providers like OpenAI are not incentivized to reduce your token costs, creating lock-in. Independent agent labs (e.g., Cognition) act like brokers, routing tasks to the most efficient models, which is a more sustainable and affordable long-term strategy for building your company.
Contrary to the idea that AI abstracts away technical details, managing AI software agents effectively requires a deeper technical understanding. You become an engineering manager who must grasp concepts like Postgres, production environments, and migrations to guide your agents and approve their work.
Running premium AI models constantly is prohibitively expensive. A cost-effective strategy is to use a cheaper model as a "manager" to understand a high-level goal, break it down, and then delegate the execution of sub-tasks to multiple, short-lived "child" sessions running more powerful models.
You can't mentally track all important customer events. Automate an agent to scan production databases daily, summarize what your paid customers did, and present it in a digestible report with links to the UI. This provides invaluable, otherwise hidden, insights into product usage and potential issues.
Developing with AI agents on your local machine creates bottlenecks and coordination nightmares. Cloud-based virtual machines (VMs) allow you to run numerous agents in parallel without code collision, drastically increasing your shipping velocity and making local development a relic of the past.
AI agents work so fast that they create a constant need for human input. Your role shifts to being a high-frequency decision-maker, requiring new systems like pinning important threads and setting 25-minute check-in cadences to avoid burnout and maintain velocity.
Don't just automate tasks; automate quality control. Create an agent that reviews a core part of your app daily, grades it against a rubric you define, and automatically spins up a new "child" agent to fix anything that scores below a certain threshold, creating a virtuous cycle of improvement.