We scan new podcasts and send you the top 5 insights daily.
Brooke Lacey downloaded her own TikTok videos to Google Drive because the platform blocks scraping. She then had Claude study the footage and comments, including a specific hook where her laugh signals the joke, so it could encode how she does hooks, cadence, and cuts.
Instead of asking AI for generic ideas, create a continuous stream of personalized inspiration. By saving posts you like to a public Instagram or TikTok folder, you can task an AI like Claude to automatically collect, analyze, and generate highly relevant content ideas tailored to your niche.
The company built an internal system that triggers when a YouTube video hits a viewership threshold. It feeds the transcript into Claude AI, which is trained to identify viral moments. These suggested clips are then sent to an editor in Descript for quick finalization, dramatically scaling short-form video production.
To master meme and slang translation, Z.AI trains models on data from public but niche online spaces like TikTok comment sections, where language is highly contextual and 'naughty.' This strategy, combined with creating synthetic data, allows their models to understand cryptic, emoji-laden communication that conventional datasets miss.
To create high-performing videos, don't invent from scratch. Find viral content in your niche and replicate its structural elements—the on-screen headline and the first few seconds of the spoken hook. Then, deliver your own unique insights within that proven format.
Brooke Lacey pairs a downloaded TikTok video with its live comment thread. She asks Claude to read the comments and propose new recording ideas from them, so the next videos directly answer real audience questions like whether hospitals will use AI for billing and coding.
Brown uses his AI "YouTube Researcher" skill to scrape the transcript of a viral video from another creator. He then prompts the AI to distill its structure into a "hook outline." This provides a proven template that he can apply to his own original content ideas.
Platforms like TikTok are releasing downloadable "AI Skills"—pre-trained instructions for AI tools. Since these skills are modifiable and run in external AIs like ChatGPT, a marketer could theoretically use TikTok’s viral video skill to generate ad concepts for Meta's platforms, breaking down ecosystem silos.
Instead of writing detailed prompts, record a screen-capture video with a tool like Supercut while verbalizing desired changes. Its AI can accurately parse long videos and map transcripts to frames, providing enough context for a model like Claude to build the feature directly with high accuracy.
Instead of asking AI for general ideas, feed it a successful piece of content and use a detailed prompt to reverse-engineer its success. Ask the AI to analyze specific components like hook type, format, pacing, editing style, and audio. This provides a reusable template you can adapt to your own niche.
To succeed on TikTok, Ladder's team obsessively analyzed their winning organic content on whiteboards, breaking down every variable: the hook, the creator's clothing, the gym setting, the specific words used. This deep qualitative analysis was crucial for understanding what the algorithm would amplify.