ChatGPT
First met in aicp-c3-basecamp-01 · Chat assistants
ChatGPTFirst met in aicp-c3-basecamp-01 · Chat assistants
OpenAI's assistant — the course's default thinking and planning surface, and the tool every mechanism gets demonstrated on.
Planning, brainstorming and turning a vague idea into a spec — the BC2 trainer's stated default for that work, even though he lives in Claude for connectors.
Reads screenshots well enough to turn competitor pricing pages into a feature map
Apps (its MCP equivalent) connect Canva, Figma and Spotify from the browser
The reference surface for the whole course: tokenization, hallucination and prompting were all demonstrated here
Plans and drafts: feature extraction, PRDs, and prompts to feed a vibe-coding platform
Runs the four-step pipeline you can inspect — tokenizer visualization, tool-use traces
Connects browser-based apps under Settings → Apps
Debugs from pasted screenshots of errors
Doesn't reach your local files — apps are browser-based, unlike Claude's desktop connectors
Doesn't run more than one tool per chat (as of BC5) — selecting image generation drops the connected app
Doesn't offer Google integrations — rivals don't ship each other's connectors
The planning and research seat at the top of a build, and the second opinion when Claude is being stubborn.
Chat assistants
The pipeline runs on the Claude stack; no second assistant subscription needed.
Relevant mainly as the surface the course teaches on — screenshots and demos in the KB will show ChatGPT even where Paul would use Claude.
current
claude: the trainer moved to Claude for connectors while keeping ChatGPT for brainstorming
chatgpt-apps: its connector layer
lovable: ChatGPT writes the prompt, Lovable builds the app
| Session | Coverage | Moment | What happened |
|---|---|---|---|
| Basecamp 1: Prompting & RAGs | demonstrated | 0:37:24 | Tokenization visualization, hallucination demo ('Claude Juggernaut'), bad vs personalized diet-plan prompting |
| Basecamp 2: Vibe Coding | demonstrated | 0:54:17 | Whole planning layer: feature extraction from pricing screenshots, MVP definition, PRD, and Lovable-ready prompt generation; trainer's current brainstorming default over Claude |
| Basecamp 5: MCP | demonstrated | 2:23:07 | Rival connector ecosystem; apps feature ~5-6 months old per trainer, reliability ~4/10 vs Claude 7/10 |
| Session 5: Mastering Lead Generation | demonstrated | 1:39:51 | The ICP research chain: capability → ranked industries → titles/size/geo/keywords → final targeting spec merge |
| Session 7: The Outreach Engine | demonstrated | 0:55:42 | Projects + Sources as the skill runtime: attach skill.md, prompt with your service, get the offer |
| Session 13: Claude Code — Skills, Sub Agents & the Token Economy (posted as 'Beyond Vibe Coding — n8n & More') | demonstrated | 1:02:17 | The discovery engine: MCP lookup, design top-10, install commands |
| Session 16: GitHub 101 — Fundamentals & Essentials | demonstrated | 1:21:05 | Escape-room ideation (20 themed variations) and the Lovable build prompt; conversation link shared with the cohort |
| Session 17: Facebook Ads Competitor Research & Replication | demonstrated | 1:09:06 | Split-screen code generator for every code node, the Apify curl, the JSON parsing, the ad-generation system prompt, and live debugging; trainer's paid subscription |
| Session 18: Nano Banana + Veo 3.1 Multi-Shot Video — UGC Genie, an AI-UGC Product Built Live | demonstrated | Desktop app; project-folder context binding ritual repeated ('drag the chat inside') | |
| Session 19: Content Strategy Pipeline — the YouTube Growth Stack, Built by an Agent Team of Chats | demonstrated | Desktop app: projects, drag-chat-into-folder context binding, terminal button shown to a learner | |
| Session 20: Complete LinkedIn Agency Automation — Carousels + Newsletter from One Content System | demonstrated | Desktop app hosting the project; Files tree explored | |
| Session 21: Advanced Data Extraction from PDFs — Parsing, OCR, and the Token-Free Product | demonstrated | Desktop app hosting the AuthReady project; folder-tree visibility gap puzzled over live | |
| Session 22: Automate Admin Tasks — Advanced RAG (Enterprise 'Internal Perplexity' Built Live) | demonstrated | Desktop app hosting the Codex project ('Talent Lens') | |
| Chapter 3: The Workflow — Four Steps, Four Agreement Rounds, No Going Back | demonstrated | His custom prompt digests the questionnaire and feeds the sitemap step | |
| Chapter 5a: The Client Build — Discovery to Direction with Lucy's Bakery | demonstrated | Role-primed sitemap cross-check; section-copy drafting with the anti-copywriter re-prompt | |
| Chapter 5c: The Finish — Details, Feedback, AI Images, and the Polished Preview | demonstrated | Copy drafting and palette-constrained image generation - 'I would even start here these days' | |
| AI Sprint: OpenClaw — Session 2 (Token Anatomy, Claude Code Inside the Server, Skills, and Multi-Agent Workspaces) | demonstrated | The patient-technical-partner starter prompt: one command at a time, screenshots back — and the live rookie-mistake catch | |
| AI Sprint: AI Evals & Reliability — Day 2 (Testing Systems That Never Answer Twice) | demonstrated | The same-prompt-twice non-determinism demo | |
| Session 1: Getting your first Client | explained | 1:01:22 | The market-validation research partner: top-5 pain prompts, niche ranking ('what companies rely on cold email?' → recruiting), TAM and gap questions |
| Practice Lab: Six Confidence-Building Projects | explained | Project 6's Custom GPT/Project home | |
| Learning Topics: Sessions 1–5 — Foundations to AI Director | explained | Named across sessions as general assistant, Custom GPT builder (paid plan required in the source snapshot), and voice-mode rehearsal surface | |
| Basecamp 3: Introduction to n8n — I | mentioned | 1:30:55 | Cloud-model roster in the chat-model node; Grok-OpenAI legal battle quip |
| Session 2: How to Plan, Scope and Sell your AI Projects | mentioned | 0:16:25 | Selfie → professional headshot; banner generation; ChatGPT as the AEO citation engine |
| Beginner Glossary (30 terms) + I'm Stuck Guides (23) | mentioned | Named in the paid-plan token guide and tool-selection guide |