# AI-Enabled Product Management and Public Goods Builder Workflows
- Session: 0f93e4bf-fff8-44e0-86a3-370981a39194
- Channel: Discord #🐣│cohort-voice
- Started: 2026-06-24T16:58:48.015Z
- Ended: 2026-06-24T17:40:37.337Z
- Participants: takekek, duckanbro, graven | Flow State, Aphilos • Pharo, ECWireless, samkuhlmann
- Tags: ai-product-management, software-development, public-goods, builder-relations, testing-and-qa, scope-creep, crypto, octant, flow-state, small-teams
## TL;DR
Duckanbro interviewed Graven about how AI has changed product management, builder relations, software delivery, and crypto/public goods work. Graven described AI as dramatically increasing iteration speed and ambition, but also shifting bottlenecks toward testing, documentation, decision-making, taste, distribution, and production readiness.
## Summary
The session was a cohort interview focused on how people are using AI in their work and how those changes inform RaidGuild's agency strategy and public content experiments. Duckanbro framed the interview series as short public conversations that will be summarized into blog posts, YouTube clips, research leads, and strategy inputs.
Graven introduced himself as an independent builder working around public goods, open source, radical markets, Superfluid-powered funding flows, and Flow State. He described himself as product-manager-like rather than a traditional developer, with deep experience working alongside developers. He recently moved into a full-time builder relations and protocol-focused role with Octant, helping grow the V2 protocol and support builders around it.
On AI tooling, Graven said Claude Code in the terminal is his main setup, often paired with GitHub Desktop. He has experimented with Codex and OpenAI, uses Claude on mobile for lightweight chatbot work, and has not adopted GitHub Copilot day to day. He found some larger-context models like Gemini less useful because they lacked the kind of context awareness he needed for real work.
The biggest shift for him is that AI has changed the speed and ambition of product development. Small teams can now ship enhancements overnight that previously would have taken a week or more. This allowed teams to say yes to more client requests, but it also created scope-creep pressure. Graven emphasized that teams must now ask not only whether they can build something, but whether they should, and what advantage the feature actually creates.
He described a major bottleneck shift: development became faster, so testing, documentation, business logic validation, and product judgment became more important. In his own work, faster shipping meant he spent much more time as the main tester, catching rough edges and gaps between the shipped product and the intended product. In response, his team started building their own evaluation harnesses, specs, feasibility checks, documentation workflows, and process scaffolding to reduce the testing and quality burden.
Graven sees AI as especially powerful for research, codebase exploration, catching up on historical context, interrogating remote GitHub repositories, and generating diagrams. He uses Excalidraw/MCP-style workflows heavily for diagramming and system understanding. He is more skeptical of AI-generated frontend/design work, noting that it often produces rough edges or generic-looking websites unless there is strong design direction.
On production readiness, Graven distinguished between AI-assisted experimentation and shipping serious systems. He is comfortable using AI to prototype, validate ideas, and make enhancements, but is still cautious with smart contracts and production-grade systems. He would want experienced developers to review AI-generated smart contract work before deployment, especially for anything beyond very simple contracts.
A recurring theme was the value of fast, sloppy validation at the low end and highly crafted production work at the high end. Graven described this as a barbell effect: AI enables cheap experiments, but once a direction proves valuable, taste, quality, decision-making, and distribution matter more than ever. He expects this to affect company structure too, creating opportunities for small, fast, aligned teams while also increasing the power of large players with existing distribution.
The conversation closed with reflections on the broader crypto market. Graven said crypto feels harder this cycle, with most excitement in tech currently centered on AI. He still sees possible stories around agents using crypto rails, but said those opportunities do not yet feel tangible at the consumer level and much current activity feels institutional. Chat participants also shared Graven's Flow State, X profile, and Octant links.
## Action Items
- Share interview outputs with Graven: Send Graven the generated summaries, clips, posts, or research outputs from this interview series once prepared. (owner: duckanbro)
- Use Graven interview themes in synthesis work: Fold Graven's themes into the broader cross-interview analysis, especially bottleneck shifts, AI-assisted product management, testing/documentation pressure, and the barbell model of experimentation versus craft. (owner: duckanbro)
- Capture Graven's public links: Include Graven's shared links as source/context for any profile or content follow-up.
## Notable Quotes
- graven | Flow State: "The bottlenecks in software development have completely changed."
- AI has made implementation faster, which shifts pressure onto testing, documentation, product judgment, and quality control.
- graven | Flow State: "Maybe yes we could, but should we?"
- Because AI makes more features possible, teams need stronger judgment about whether each feature is actually worth building.
- graven | Flow State: "Development is a bottleneck and it pushes it to other parts, like distribution becomes more and more valuable."
- As AI reduces the cost of building, advantages move toward distribution, customer access, taste, and decision-making.
- graven | Flow State: "You can do a sloppy first version as an intro and demo validation and then work from there."
- AI is highly useful for quick prototypes and validation before deciding whether an idea deserves production investment.
- graven | Flow State: "I haven't gotten to the level where I would say, yeah, apply code with a smart contract and let's roll it out."
- He remains cautious about deploying AI-generated smart contracts without professional developer review.