Article · 2026-07-20 · 6 moments

🎙️ How I AI: How the founder of Morning Brew built a Claude content machine that never runs out of ideas

Your weekly listens from How I AI, part of the Lenny’s Podcast Network ✦ AI generated

Mechanism◆

A daily AI 'Oracle' that scans Slack, Notion, Gmail, and meeting notes to surface ranked content ideas eliminates the blank page, since finding what to say—not writing—is what actually stalls most creators.

Alex Lieberman built an 'Oracle' that scans a week of Slack, Notion, Gmail, and meeting notes plus followed X/LinkedIn accounts to surface 15 ranked content ideas daily, arguing that idea generation—not writing—is the real bottleneck.

transcript

Alex Lieberman: The blank page is the real enemy of consistent content, and an AI Oracle makes it disappear. Alex's Oracle scans seven days of Slack, Notion, Gmail, and meeting notes, then surfaces 15 ranked content spikes each day—half from internal sources, half from the X and LinkedIn accounts he follows.

Claim◆

AI-generated content slop happens when the human interview subject fails to share interesting ideas, not because the underlying model is deficient.

Lieberman argues that his Content Machine only produces 'slop' when the person being interviewed doesn't provide interesting material, since the model shapes ideas rather than inventing them.

transcript

Alex Lieberman: AI slop is mostly a people problem, not a model problem. Alex's framing here is blunt: the only time the Content Machine produces slop is when the person being interviewed doesn't share interesting enough ideas during the interview. The model is shaping clay, not inventing ideas.

rebuts · 1

Mechanism

A 'Writer's Council' of six AI writer personas scores every draft and forces revision loops until it clears a 9-out-of-10 threshold, removing the need for Alex to manually police every post.

Six writer personas, including one Alex calls the 'AI slop allergist,' each score drafts and trigger revisions until the aggregate score clears a 9/10 bar, letting the system hold quality standards automatically.

transcript

Alex Lieberman: Six writer personas (including a character Alex calls "the AI slop allergist") each score the draft and run a revision loop until the aggregate clears the threshold. That scoring mechanism is also what keeps Alex from having to manually police every post, so the system holds its standard even when he's in a hurry.

Mechanism

Comparing each draft to its published version and extracting generalizable lessons into a permanent file creates a reinforcement loop that makes the AI system stop repeating mistakes over time.

After every piece, Lieberman's system compares the draft to what was actually published, extracts lessons, and asks whether to save them permanently—building a compounding feedback loop few people bother to create.

transcript

Alex Lieberman: After every piece, Alex's Content Machine compares the original draft to the published version, extracts generalizable lessons, and asks if they should be added to a permanent lessons file. Over time, the system stops making the same mistakes. This is the kind of feedback loop most people never build.

Mechanism

Deploying six AI interviewer personas modeled on real interviewers to ask follow-up questions until enough specifics are extracted is the most-skipped but most important step in AI content creation.

The Content Machine runs six interviewer personas (Tim Ferriss, Joe Rogan, Larry King, Howard Stern, Barbara Walters, Michael Barrow) that push for specifics, with Lieberman answering by voice via Wispr Flow so every line traces back to something he actually said.

transcript

Alex Lieberman: The system deploys six interviewer personas (Tim Ferriss, Joe Rogan, Larry King, Howard Stern, Barbara Walters, Michael Barrow) that ask follow-up questions until they've extracted enough specifics to write something real. Alex voice-to-texts his answers via Wispr Flow, so every sentence in the final draft traces back to something he actually said.

Mechanism

Drafting AI content against a personal Markdown 'voice guide' of top-performing posts, hooks, and language patterns produces writing calibrated to the individual rather than to a generic internet average.

Lieberman's Content Machine drafts from a Markdown file capturing his best posts, hook formulas, and language patterns like 'self-deprecating confidence,' keeping output calibrated to his actual voice.

transcript

Alex Lieberman: Alex's Content Machine pulls from a personal voice guide that captures his top-performing posts, hook formulas, content structures, and even specific language patterns like "self-deprecating confidence." The system drafts against that file, which means it's calibrated to how Alex actually writes rather than to some average of the internet.

supports · 1

Highlight slides

The Blank Page Isn't the Problem✦ from: A daily AI 'Oracle' that scans Slack, Notion, Gmail, and meeting notes to surface ranked content ideas eliminates the blank page, since finding what to say—not writing—is what actually stalls most creators.AI Slop Is a People Problem✦ from: AI-generated content slop happens when the human interview subject fails to share interesting ideas, not because the underlying model is deficient.How the Oracle Works✦ from: A daily AI 'Oracle' that scans Slack, Notion, Gmail, and meeting notes to surface ranked content ideas eliminates the blank page, since finding what to say—not writing—is what actually stalls most creators.The Model Shapes, It Doesn't Invent✦ from: AI-generated content slop happens when the human interview subject fails to share interesting ideas, not because the underlying model is deficient.Where Ideas Come From✦ from: A daily AI 'Oracle' that scans Slack, Notion, Gmail, and meeting notes to surface ranked content ideas eliminates the blank page, since finding what to say—not writing—is what actually stalls most creators.

Related episodes

Article · Simon Willison's Weblog

Claude make Fable 5 permanent

→

Video · The Cognitive Revolution

Alignment with Awakening: Davidad on Moral Realism, AI Wisdom, & why His p(Doom) is Down to 5%

→

Video · The Cognitive Revolution

Fable's Back, AI Engineer Recap, & SambaNova

→