Ruby Central, Vibe Coding Ceilings, and What Still Requires a Human: Michael Rispoli on the Work AI Cannot Take
The floor is dropping out from under a lot of things that used to feel stable. And the people still standing are the ones who never stopped doing the hard part.
The floor is dropping out from under a lot of things that used to feel stable in this industry, and this episode is an honest look at what that actually means. We sit down with Michael Rispoli, co-founder and CTO of Cause of a Kind, to work through what is happening to Ruby Central, what the Gusto announcement really signals, and whether the structure the Ruby community has depended on for years is still built for this moment. We look at what governance models have actually worked for languages over time, why a benevolent dictator gets things done that a committee never will, and what the slow fracturing you are starting to see in Ruby has in common with what already happened in JavaScript. Nobody is declaring anything over. But the conversation makes clear that the next chapter is not going to write itself.
🎧 Ruby Central, Vibe Coding Ceilings, and What Still Requires a Human: Mike Rispoli on the Work AI Cannot Take
Available now wherever you get your podcasts, including Apple Podcasts and Spotify.
One of the sharpest threads in this conversation is what happens to agency work when the floor drops out from under everything that used to be easy to sell. Cause of a Kind just rebuilt around modernization, healthcare, education, PE-backed companies cleaning up before an acquisition, because those are the places where the cost of being wrong is real and where no one is signing off on handing the job to an agent and walking away. In this episode, Mike makes it clear that the clients arriving with Lovable prototypes are not the problem. The problem is those prototypes have twenty features, no clear entry point, and no answer to what makes a stranger stop scrolling. Risk is what people pay to mitigate, and the agencies repositioning around that right now are the ones that will still have work when the easy stuff is fully gone.
What makes this episode stand out is how honest the conversation gets around staying sharp when the models are doing most of the lifting. Depending on one model turns out to be its own kind of risk, and in this conversation Mike walks through exactly how he built a multi-model workflow before the Opus pricing shock even hit. We go deep on why each model behaves differently, why low thinking on GPT-5.5 gets you closer to how you would have solved it yourself than high thinking on Claude ever did, and why knowing that difference is the skill nobody is deliberately building right now. There is a pull request in here that works but has a smell to it, and the answer is not to override it or accept it. It is to go on a walk, forget what the model did, and come back knowing exactly what to ask for.
Listen Now
Show Notes
00:00 Welcome and Guest Intro
00:32 Gusto Joins Ruby Alliance
01:59 Is Ruby Central Ending
03:28 Governance Models Debate
06:46 Conferences and Community Shift
08:15 Centralized Packages vs Git URLs
09:17 Cause of a Kind Rebrand
11:21 Modernization and War Room
14:05 AI Pressure and Agency Strategy
18:25 AI Builds Faster Rails Rewrites
21:40 Security Chaos Monkey Era
25:31 Tooling Diversification
26:50 Testing Alternatives
28:00 GPT 5.5 Workflow
30:48 Switching Model Harnesses
31:50 When to Go Solo
34:28 Vibe Coding Pitfalls
36:16 Marketing First MVP
39:02 Teaching FDE Skills
43:08 Selling Pushback
47:44 Closing and Meetup
Who Else Should We Talk To?
Know someone doing innovative or inspiring work with Ruby and AI? Let us know we might feature them in an upcoming episode. Contact us at news@therubyaipodcast.com.
Let’s continue to redefine what’s possible with Ruby and AI!
💎 Valentino & Joe 💎
 | A guest post by
|