Automation at the speed of Swamp (Friends)
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “Automation at the speed of Swamp (Friends)” inside PodZeus.
In this three-part episode of The Changelog, host Adam and guest Adam Jacob explore the revolutionary impact of AI-driven automation through Swamp, a system that uses agentic AI to build, manage, and scale software infrastructure workflows at unprecedented speed. The conversation reveals a paradigm shift in software development: AI is no longer a tool but a co-developer that constructs the very systems responsible for building software, rendering traditional development and operations workflows obsolete. Swamp enables teams to define high-level architectural intent in natural language, which AI agents then translate into self-validating, reusable automation—handling everything from provisioning VMs to debugging network issues—while eliminating the need for stand-ups, manual code reviews, and testing. The episode features a live demo showcasing Swamp’s ability to inventory Proxmox VMs, delete instances, and create workflows in minutes, illustrating how the platform collapses silos across infrastructure, deployment, and monitoring into a unified, autonomous layer of execution. Despite the emotional disorientation and professional upheaval this transformation brings, the hosts express deep optimism about the future of engineering, where the most valuable skill is not coding, but architectural vision and intent expression. The discussion delves into the broader implications of this shift: the potential obsolescence of traditional engineering roles, the ethical challenges of open-source contributions in an agent-driven world, and the existential weight of leading teams through rapid, disruptive change. Yet, the sentiment remains resolutely positive, emphasizing that those who adapt early—by relearning how to guide AI agents and think at a higher level of abstraction—will become indispensable in shaping the next generation of software systems. The episode closes with a personal reflection on the emotional cost of innovation, the pride in Adam Jacob’s journey, and a passionate call to action for listeners to join the Swamp community, explore its codebase, and contribute to a future where software development is not about writing code, but about defining outcomes and enabling AI to build them at the speed of thought.
AI is no longer a tool but a co-developer that builds the systems responsible for software creation, fundamentally changing the development lifecycle.
Traditional software roles are evolving—engineers must now focus on architecture and intent expression, not manual coding, to remain relevant.
Swamp enables agentic automation that collapses infrastructure, deployment, and monitoring into unified, self-validating workflows executed in minutes.
The future of software development is AI-native: teams define outcomes, and AI handles implementation, drastically reducing development cycles from days to minutes.
While the transition is emotionally and professionally disruptive, early adopters who master architectural thinking will lead the next era of innovation.
…and 2 more takeaways available in PodZeus
The AI Revolution: From Tool to Co-Builder
“I don't know. Like, like if, if I had a hard, what's a, what's a hard day at work look like when what I'm doing is telling agents to do the work for me and then the system is assembling it.”
Swamp: The System That Builds Itself
“It's like, hey, I just told it to like write me a workflow to reboot Procmox. And they're like, I know. So you wrote a script and you're like, no, no, I really did it. I really didn't.”
The Death of the Traditional Developer Workflow
The hosts discuss how traditional development practices—like stand-ups, code reviews, and manual testing—are no longer viable at the speed of AI-driven development. Instead, they describe a new workflow where adversarial agents handle code validation, UAT testing, and even debugging. The team uses a 'factory pattern' to generate multiple instances of infrastructure, and all data is stored and queried by the system.
The Collapse of Traditional Workflows
“You could just tell it, go look at environment A and B. Tell me what's different and write a workflow to fix them.”
From Toy to Production: Trusting the Agent
The episode confronts the skepticism around using AI for production systems. Adam argues that transparency—being able to inspect every action, credential, and workflow—makes agent-driven systems more trustworthy than legacy, opaque CI/CD pipelines.
“The only boundary of trust I have is the inputs of a human being at the very top of the funnel.”
“I can already see how this changes the entire game for what I was trying to do. And that's kind of interesting now because if Swamp is a critical component to this thing I build, well, there you go. There's the toll booth for you.”
“It's like, hey, I just told it to like write me a workflow to reboot Procmox. And they're like, I know. So you wrote a script and you're like, no, no, I really did it. I really didn't.”
Hosts
Guest
Swamp
other
Adam Jacob
person
Proxmox
product
Swamp Club
product
Domain-Driven Design
book
Coder.com
organization
Tailscale
organization
Postgres
product
Block
organization
RWX
organization
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “Automation at the speed of Swamp (Friends)” inside PodZeus.
Start discovering podcast insights today
Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.
No credit card required • 7-day trial • Cancel anytime