About this video

• Video Title: Screensharing How This Guy Makes Millions with AI Agents
• Channel: Greg Isenberg
• Speakers: Greg Isenberg, Rowan Chung
• Duration: 00:43:07

Overview

This video features a conversation between Greg Isenberg and Rowan Chung, founder of The Rundown AI. Rowan shares seven specific ways he leverages AI agents to streamline business operations and content creation, enabling his relatively small team to function like a much larger one. The discussion covers practical applications of AI, including content generation, customer service, and internal processes, with a focus on providing actionable insights and tools.

Key takeaways

• AI-powered content creation: Utilizing AI avatars (using tools like HeyGen and 11 Labs) to generate short-form videos from written content, achieving significant follower growth.
• Efficient tweet generation: Training an AI model on past top-performing tweets to generate new content, which is then refined with human insight.
• AI as a newsletter editor: Employing AI (like Claude) as an editor-in-chief to refine newsletter content, with strategies to overcome length limitations in AI conversations.
• Automated meeting scheduling: Using an AI agent (Lindy) to manage meeting schedules, reducing the need for manual coordination and avoiding the perceived disrespect of sending scheduling links.
• AI for research and learning: Leveraging AI tools like Perplexity Comet to summarize videos and learn complex topics, acting as a personalized AI tutor.
• Streamlined new hire onboarding: Implementing a custom GPT trained on company SOPs and documents to answer new hires' questions, freeing up manager time.
• Email management and lead qualification: Using AI agents to filter customer support inboxes, identify key feedback, ping relevant team members, and draft replies, even landing sponsorships.

Rowan Chung trains an AI to generate tweets in his writing style by:

1. Collecting top-performing tweets: He gathers 10 to 20 of his best-performing tweets.
2. Using a prompt with Claude: He inputs these tweets into Claude along with a specific prompt designed to train the AI on his writing style.
3. Voice dictation workflow: He uses WhisperFlow to dictate his ideas while walking, creating voice notes.
4. Pushing notes into the trained chat: He then inputs these voice notes into the Claude chat that has been trained on his tweet writing style.

This process results in tweets that are approximately 90% ready, requiring minimal human editing for final touches and added personal insights.