Admin Admin Podcast Admin Admin Podcast #102 – Getting the band back together
Nov 21, 2025
After a break, the hosts reunite to share life updates and dive into the fascinating world of large language models (LLMs). They discuss personal experiences using tools like ChatGPT and concerns surrounding data privacy. Practical applications emerge, with tips for generating scripts and reports using AI. The conversation touches on ethical implications and the impact on skills development for junior engineers. As they ponder the balance of trust in LLM-generated code, the hosts also brainstorm plans for their future discussions.
AI Snips
Chapters
Transcript
Episode notes
LLMs Are A Double-Edged Tool
- LLMs have quickly become central and controversial in ops and security work.
- They present huge productivity gains but also significant data-privacy and hallucination risks.
ChatGPT Helped Build A Postgres Backup Job
- Gerry used ChatGPT to generate a Python script to dump Postgres tables to an S3 bucket and adapted issues using his own AWS knowledge.
- He discovered Kubernetes Jobs and CronJobs as a simpler alternative to Lambda for this task.
Use LLMs For Drafting And Reviewing, But Verify
- Use code assistants like Copilot for grammar, PR explanations, and framing communications.
- Always verify outputs and adjust tone or correctness before sharing or committing.