
Sign up to save your podcasts
Or


If you ever hear someone yell “WHAT SUCKS ABOUT YOUR JOB?” at a meetup, there’s a good chance it’s us.
In this episode, Megan and Nicole finally sit down after a long podcast break to debrief two big AI Portland events: an October session with David Long, VP of Digital and Innovation for the Portland Trail Blazers, and a very nerdy, very packed November deep dive into MCP (Model Context Protocol). Along the way, there’s a birthday, some early-morning chaos, and a few strong opinions about agents, hype, and where AI very much does not belong.
They dig into how the Blazers are actually using AI (and when they deliberately don’t), why “what sucks about your job?” is a surprisingly powerful innovation question, and what MCP really is for the non-engineers in the back. Then they recap live demos from DevSwarm, shout out Radek from Keboola for the emergency hero fill-in, and talk about why good engineering fundamentals matter more than ever when you give the robots access to your codebase.
Along the way:
🏀 How the Portland Trail Blazers are using AI to solve real problems, not invented ones
🧠 The Einstellung effect and why “this is just how we do it” is killing innovation
🛠️ MCP 101: what it is, why it matters, and why even non-devs were taking notes
🧩 DevSwarm’s demos: from Confluence specs to JIRA tickets to Figma-to-code flows
📏 Why human-in-the-loop, architecture, style guides, and documentation matter _more_ with AI, not less
Plus: Stacklok’s token-saving magic for MCP integrations, the never-ending trough of disillusionment, and why 2025 still feels like “the year of trying to actually get productive with AI.”
Related links:
Special thanks to our speakers:
And to our event sponsors:
Chapters
00:00 Introduction and Personal Reflections
02:50 AI Portland October Event Recap
04:02 Innovation in Organizations
10:54 AI Portland November Event Recap (MCP)
21:23 Closing Thoughts and Future Directions
25:48 Stacklok Token Optimization
By AI Portland4.7
77 ratings
If you ever hear someone yell “WHAT SUCKS ABOUT YOUR JOB?” at a meetup, there’s a good chance it’s us.
In this episode, Megan and Nicole finally sit down after a long podcast break to debrief two big AI Portland events: an October session with David Long, VP of Digital and Innovation for the Portland Trail Blazers, and a very nerdy, very packed November deep dive into MCP (Model Context Protocol). Along the way, there’s a birthday, some early-morning chaos, and a few strong opinions about agents, hype, and where AI very much does not belong.
They dig into how the Blazers are actually using AI (and when they deliberately don’t), why “what sucks about your job?” is a surprisingly powerful innovation question, and what MCP really is for the non-engineers in the back. Then they recap live demos from DevSwarm, shout out Radek from Keboola for the emergency hero fill-in, and talk about why good engineering fundamentals matter more than ever when you give the robots access to your codebase.
Along the way:
🏀 How the Portland Trail Blazers are using AI to solve real problems, not invented ones
🧠 The Einstellung effect and why “this is just how we do it” is killing innovation
🛠️ MCP 101: what it is, why it matters, and why even non-devs were taking notes
🧩 DevSwarm’s demos: from Confluence specs to JIRA tickets to Figma-to-code flows
📏 Why human-in-the-loop, architecture, style guides, and documentation matter _more_ with AI, not less
Plus: Stacklok’s token-saving magic for MCP integrations, the never-ending trough of disillusionment, and why 2025 still feels like “the year of trying to actually get productive with AI.”
Related links:
Special thanks to our speakers:
And to our event sponsors:
Chapters
00:00 Introduction and Personal Reflections
02:50 AI Portland October Event Recap
04:02 Innovation in Organizations
10:54 AI Portland November Event Recap (MCP)
21:23 Closing Thoughts and Future Directions
25:48 Stacklok Token Optimization