# Episode 809 | What I Learned Diving into A.I. for 100 Days (with Craig Hewitt)

Page: https://stenobird.com/podcast/startups-for-the-rest-of-us/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt
Text version: https://stenobird.com/podcast/startups-for-the-rest-of-us/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt.md
Podcast: [Startups For the Rest of Us](https://stenobird.com/podcast/startups-for-the-rest-of-us)
Published: 2025-12-02T10:00:30+00:00
Episode link: https://www.startupsfortherestofus.com/episodes/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt
Audio file: https://prfx.byspotify.com/e/episodes.castos.com/5e1c969972fbb8-96341148/2217990/c1e-rgx1fw424wu7qj60-6zqn6qvvbo0k-ath9ib.mp3
Processing state: processed
JSON: https://stenobird.com/v1/public/podcasts/startups-for-the-rest-of-us/episodes/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt
Duration seconds: 2371

## Resource

Craig Hewitt shares the practical results of his 100-day experiment testing various AI tools for SaaS founders. The discussion highlights specific high-value automations and why general-purpose LLMs might not be the best choice for technical workflows.

## Highlights
- Main idea: Specialized AI agents, like DocsBot, can significantly reduce customer support overhead by leveraging existing knowledge bases
- Practical takeaway: Use 'escape hatches' in AI implementations to ensure users can easily transition to human support if the bot fails
- Failure mode: Relying solely on general-purpose tools like ChatGPT for complex technical tasks can be less efficient than using specialized coding assistants
- Main idea: AI-driven automation in content production, such as using Creator Hooks, can streamline video and social media workflows
- Practical takeaway: When evaluating a second product idea, prioritize personal excitement and the potential to leverage existing customer bases

## Topics

SaaS, Artificial Intelligence, AI Agents, Customer Support Automation, Product Development, Content Automation, Entrepreneurship, LLMs

## Chapters
- 1:00 — The 100 Days of AI Experiment: An introduction to Craig Hewitt's deep dive into the evolving landscape of AI tools for entrepreneurs.
- 6:40 — Best-in-Class AI Tool Categories: A breakdown of the most effective tools for coding, general purpose chat, and automation.
- 9:40 — Custom GPTs and Reusable Workflows: The benefits of creating specialized GPTs trained on specific datasets for team use.
- 15:20 — The Rise of Agentic Automation: Evaluating the current utility and future potential of AI agents in business workflows.
- 18:10 — Reducing Support Burden with AI Agents: How implementing a vector-store-based chatbot halved customer support volume at Castos.
- 24:20 — AI for Content and Video Production: Using tools like Creator Hooks to automate title and thumbnail generation for YouTube.
- 30:10 — Evaluating New SaaS Product Ideas: A look at Craig's new project, LinkBerry.ai, and the emotional challenges of launching a second product.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/startups-for-the-rest-of-us/episodes/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/startups-for-the-rest-of-us/episode-809-what-i-learned-diving-into-a-i-for-100-days-with-craig-hewitt.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.