Get AI summaries of any video or article — Sign up free
Generate 42+ Viral Content Ideas In Seconds With Kortex AI thumbnail

Generate 42+ Viral Content Ideas In Seconds With Kortex AI

Noah Vincent·
5 min read

Based on Noah Vincent's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Use the “2-year-old test” interview prompt to extract 20 content titles from concrete life experiences across health, relationships, finances, creativity, and challenges.

Briefing

A set of three Cortex AI prompts turns personal experience and audience research into large batches of ready-to-write content ideas—often hundreds of angles—from a single input. The core payoff is speed plus specificity: instead of staring at a blank page, creators can generate interview-based title lists, angle variations for one subject, and fully categorized “ID” (idea) outputs that map to different content styles.

The first prompt, dubbed the “2-year-old test,” converts Cortex into an interviewer. It runs a structured interview about the creator’s life over the past two to five years—covering actions taken to expand knowledge, improve health, strengthen relationships, boost financial stability, broaden perspective through new experiences, push comfort zones, develop creativity, and overcome challenges. After the interview, Cortex produces a list of 20 content titles grounded in those answers. The transcript emphasizes that more detail improves results, and it recommends using Cortex’s transcription feature: record voice responses, transcribe them into text, and feed richer context into the interview. The example answers include quitting cigarettes and caffeine, building a learning system, studying psychology and mental models, starting and restarting businesses, and documenting personal growth on YouTube—then translating those specifics into titles like content about addiction recovery, gym mindset, and investing in oneself.

The second prompt focuses on “100 angles with one subject,” turning a single topic into a branching idea engine. It instructs Cortex to act as a brainstorming assistant that generates subtopics and then questions for each subtopic, using quantifiable, proven approaches. The prompt also forces action-oriented framing by requiring each subtopic to start with a verb, while explicitly avoiding overly specific tactics that would lock the output into narrow techniques. The workflow advantage is that Cortex can pull in documents as context—such as an ICP (ideal customer profile) or persona notes—so the AI iterates based on existing workspace material rather than inventing from scratch. In the demonstration, a broad theme about building a business for self-realization becomes multiple subtopics (e.g., aligning a business with values, refining skills, establishing brand identity), and each subtopic yields multiple question prompts that can be turned into newsletters, YouTube videos, or other formats. The same mechanism is shown again with a different subject (modern food and the “prime diet” framing), producing additional angle sets like skills, mistakes, benefits, steps, and success stories.

The final prompt is an “ID generator” designed to produce categorized content ideas. It splits outputs into four ID types—practical, analytical, aspirational, and anthropological (the transcript uses “anthropological” language)—and further organizes them by subthemes such as tips/tools, trends/numbers/reasoning, lessons/errors/personal stories, and failures/struggles/paradoxes/observations. With only a topic description, an audience, and a desired outcome, Cortex generates one idea per subtopic category. The example results include titles like “Five simple shifts to align your business with your source purpose” (practical), “The rise of purpose-driven business” (analytical), and “From burnout to bliss” (aspirational), plus value-measurement concepts like a “freedom score card.”

Taken together, the prompts are positioned as a repeatable system: mine lived experience, expand one topic into many action-based angles, and generate categorized titles that match different audience motivations—so content production becomes a pipeline rather than a daily struggle.

Cornell Notes

Cortex AI can generate large batches of content ideas by using three structured prompts. First, the “2-year-old test” runs an interview about a creator’s past two to five years and then outputs 20 titles based on those answers, with better results when responses are detailed (including voice transcription). Second, “100 angles with one subject” turns one topic into multiple verb-led subtopics and then question prompts tied to quantifiable, proven approaches, using audience documents as context (like an ICP). Third, the “ID generator” produces categorized ideas across practical, analytical, aspirational, and anthropological styles, using only a topic, audience, and outcome. This matters because it converts personal experience and audience research into ready-to-write titles and angles quickly.

How does the “2-year-old test” prompt turn personal history into publishable content titles?

It first turns Cortex into an interviewer that asks a creator to reflect on the past two to five years across specific domains: expanding knowledge/skills, improving health, strengthening relationships, improving financial stability, gaining perspective through new experiences, pushing comfort zones, boosting creativity, and overcoming challenges. After the interview, Cortex compiles 20 content titles derived from the answers. The transcript highlights that richer answers improve output and recommends using Cortex’s transcribe feature—record voice responses, transcribe them into text, and feed that detailed context into the interview prompt.

What makes “100 angles with one subject” different from simply asking for more ideas?

It forces structure and actionability. The prompt instructs Cortex to generate subtopics that start with verbs (e.g., “Design,” “Refine,” “Maximize,” “Establish”), and it requires subtopics to be results-oriented—aimed at helping readers learn a skill, implement a strategy, or solve a problem. It also includes guardrails: avoid specific tactics/techniques in subtopics so the output stays flexible and broadly useful. Then it generates questions for each subtopic using quantifiable and proven approaches, producing many distinct angles from one topic.

How does using audience documents as context improve idea generation in Cortex?

The transcript emphasizes that Cortex can use workspace documents as context, such as an ICP or persona notes. Since AI can struggle to create from scratch, feeding existing audience data helps it iterate more accurately. In practice, the prompt tells Cortex to tailor ideas to the audience’s interests, and the creator can add documents directly as context inside Cortex so the generated subtopics and questions better match the target readership.

What are the four “ID” categories, and how do they shape the type of titles produced?

The ID generator splits outputs into four types: practical, analytical, aspirational, and anthropological. Practical ideas focus on tips/tools/advice; analytical ideas emphasize trends, numbers, reasoning, examples, and debunking; aspirational ideas lean into lessons, errors, thoughts, and personal stories; anthropological ideas focus on fears, failures, struggles, paradoxes, and observations. The prompt then produces one idea per subtheme category, yielding a mix of title styles that can appeal to different audience motivations.

Why does the transcript recommend transcription (voice-to-text) during the interview prompt?

Because it increases the amount and richness of data fed into Cortex. The creator can record answers by speaking, use the transcribe button to convert speech into text, and then provide that detailed narrative context to the AI. The transcript’s logic is straightforward: more context usually leads to more specific and relevant content titles.

How can one subject realistically lead to “virtually infinite” content angles in this system?

One subject first becomes multiple subtopics (seven in the demonstration). Each subtopic then generates multiple questions (ten in the demonstration) that can be turned into content pieces. That means the number of potential outputs multiplies quickly: one topic → several subtopics → many question prompts → many distinct titles and formats. The transcript illustrates this with examples like personal branding and also with a different subject area related to modern food.

Review Questions

  1. If a creator wants more personal, emotionally grounded content titles, which prompt should they start with and what input should they maximize?
  2. How does the “100 angles with one subject” prompt enforce actionability and avoid overly narrow tactics?
  3. When using the ID generator, how would you decide which ID category to prioritize for a specific audience goal?

Key Points

  1. 1

    Use the “2-year-old test” interview prompt to extract 20 content titles from concrete life experiences across health, relationships, finances, creativity, and challenges.

  2. 2

    Improve title relevance by answering with detail and using Cortex’s transcription feature to convert voice responses into text context.

  3. 3

    Apply “100 angles with one subject” to generate verb-led subtopics and question prompts that map to results-oriented, proven approaches.

  4. 4

    Feed audience research into Cortex as context (such as an ICP document) so idea generation iterates on existing persona data rather than generic assumptions.

  5. 5

    Use the ID generator to produce categorized ideas across practical, analytical, aspirational, and anthropological styles, giving a balanced mix of title angles.

  6. 6

    Treat the prompts as a repeatable pipeline: interview → angle expansion → categorized title generation—so content output becomes systematic instead of ad hoc.

Highlights

The “2-year-old test” converts a structured two-to-five-year life interview into a list of 20 titles grounded in the creator’s own specifics.
“100 angles with one subject” multiplies output by turning one topic into verb-based subtopics and then question prompts tied to quantifiable, proven approaches.
The ID generator produces titles across four distinct content styles—practical, analytical, aspirational, and anthropological—so the same business theme can yield varied hooks.

Topics

Mentioned