The Telos Method Explained (ft. Daniel Miessler)

Core idea: “Telos” as a personal/organizational system for AI-assisted clarity

The video frames Telos/Tilos as a way to build “deep context” about yourself (or an organization) so AI can help you plan, decide, and execute with traceability from:

• day-to-day actions →
• projects →
• strategies →
• challenges →
• goals →
• mission

Telos is positioned as both:

• a self-discovery / meaning-making framework (especially amid AI-driven job disruption), and
• an AI-operational workflow, where the Telos document becomes a knowledge source that downstream tools/agents can query.

---

Practical product/features: what the Telos “file” contains

A Telos document is described as being structured (in Markdown, optionally with visuals) around these components:

1. Problem

   • The root problem you’re trying to solve
   • Example: “not enough people understand how to build secure open-source software”

2. Mission

   • The “primary mission” (usually 1; sometimes 2–3) flowing from the problem

3. Narratives

   • “Quick handles” for how you explain who you are and what you stand for
   • Also used to reduce self-doubt by reminding you of your identity

4. Goals

   • Concrete targets
   • Optionally includes “urgency vs importance” via sub-structure (e.g., “start this later”)

5. Challenges

   • What specifically blocks you from achieving the goals
   • Example themes: motivation problems, UX difficulty, social uncertainty

6. Strategies (C1/C2/C3 style ordering)

   • Specific ways you address each challenge
   • Projects tie into strategies

7. Projects

   • The concrete deliverables/actions
   • Explicitly linked to strategies

8. Additional context (optional but recommended)

   • Metrics, journal entries, predictions
   • Favorite books/movies
   • “Things I’ve been wrong about”
   • Anything that improves AI analysis

Key implementation principle

• Everything meaningful should map into a goal → strategy → project chain.
• Tasks can exist, but “meaningful long-term work” should become projects.

---

AI integration: explainability and automation

Telos’s structure is designed so AI can understand why you do what you do.

Context size as a critical constraint

• The speaker emphasizes context size as important.
• Mentions modern model context windows such as Google’s ~1M tokens input (with large output limits).

Telos + RAG

• Telos is primarily “pure text context.”
• The ecosystem also uses RAG (retrieval-augmented generation) for other content—such as transcripts.

---

Review/tutorial content: a mini “how-to build Telos”

The video includes a walkthrough:

• Start from the template and fill in:
  • Problem → Mission → Narratives → Goals → Challenges → Strategies → Projects
• Use an example Telos for secure open-source software, including narratives such as low-energy self-doubt scenarios.
• Use AI to support tasks like:
  • extracting narratives,
  • reminding you of your “most important sentence,”
  • generating later outputs (opening sentences, metric tracking prompts, dashboards, etc.).

---

Automation stack (beyond Telos): “augmentations” as an agent toolchain

A major portion describes building an augmentation stack—AI services that Telos sits on top of.

Tools/services architecture described

• AWS Bedrock agents (explicitly mentioned as an agent platform for production-ish workflows)
• AWS Step Functions for orchestrating workflows
• AWS Lambda as callable services/APIs
• n8n for workflow automation (described as fast and agent-workflow friendly)
• “Fabric”-style prompting patterns (the speaker references “Fabric” patterns and GitHub-backed templates/patterns)
• RAG via vector databases for searching transcripts/content (especially useful for video/transcript analysis)

“Pipeable services” + agent tool calling

Services are described as composable pipelines, such as:

• one tool fetches/structures data,
• another queries Telos,
• another summarizes,
• another looks up content via RAG,
• and an agent decides which tools to call.

Example automation for security work

Scenario: monitor and assess bug bounty opportunities

• Agent runs recon (e.g., parsing subdomains, reverse discovery)
• Generates security summary reports
• Sends results via email/other channels
• Optionally notifies if a bounty payout occurs

---

Local vs cloud: privacy/control concerns (tech analysis)

The speakers prefer local execution where possible, citing concerns that cloud providers may:

• remove user choice and
• force data into the cloud (example: Alexa/data behavior)

Additional concerns include:

• pipelines breaking due to third-party changes
• safety/privacy risks of browser/computer agents
  • e.g., they avoid letting agents upload friends’ data to third-party tools

---

Content/creator workflow automation (threshold-style “label and rate”)

A related tutorial describes operations for creators:

• A system collects inputs from multiple sources:
  • RSS, feeds, notes, transcripts
• A “label and rate” prompt produces JSON scoring (whether something is worth turning into a video)
• High-scoring candidates route into a working doc/list

Notes and extensions:

• Direct YouTube scraping is difficult (platform restrictions)
• Automation via RSS/polling or creator-triggered Zapier-like workflows if allowed
• Extensions include:
  • predicting resonance for video transcripts (based on past performance)
  • evaluating thumbnails (why some performed well)

---

“Telos workshops” / sessions as a guided review process

The speaker runs 2-hour Telos sessions with friends:

• walk through the person’s Telos
• extract ideas and narratives
• convert them into the framework

Reported outcomes:

• After filling it in, people often feel like they immediately know what they “should be doing.”
• The process is framed as “therapist-like”: extraction + structuring.

---

“Human 3.0” / existential labor analysis: why Telos exists

The video argues that AI may make many conventional jobs obsolete or reduce headcount (with an ideal employee count trending toward zero). This creates:

• mass anxiety/dread about relevance and fulfillment

The proposed antidote:

• self-capture (knowing what you’re about)
• using AI to amplify you via an augmentation stack

A debate is included about whether everyone will become:

• a “conference speaker/idea originator,” versus
• micro-niche broadcasting

One view suggests AI/digital assistants will “crawl” content and personalize discovery, elevating niche experts.

---

Key references & sources mentioned

• David Graeber: Bullshit Jobs (context for hollow jobs)
• Mint and Quicken Simplify (personal finance tooling examples)
• GitHub project/repo: Telos described as an open-source framework with “two files” (plus a more complex personal Telos example)
• AWS Bedrock, AWS Step Functions, AWS Lambda
• n8n
• LangChain / LangGraph
  • LangGraph recommended as an upgraded agent workflow approach
• CrewAI
  • tested but considered too hard to keep predictable/pathed
• MCP (Model Context Protocol)
  • turning apps into API-like objects/services for agents
• Zap-based webhooks / Zapier
• Index cards + notebook capture system (capture-first workflow)
• Threshold
  • described as a capture/rating system for content sources

---

Main speakers/sources (as presented in the video)

• Daniel Miessler (featured guest / source of Telos overview)
• Jason (recurring figure mentioned as collaborating on AI/agents; not the main speaker in this clip)
• Network Chuck (host/interviewer; asks most product/workflow/implementation questions)