In April, Kevin Rose and I hosted a two-day AI Accelerator in Venice Beach.

Our goal was simple: no broad abstract conversations. We wanted hands on keyboards, live demos, things breaking, and a lot of interruptions that sounded like, “Wait, show me exactly how you did that.”

Forty founders and builders spent two days working together and learning from each other. Think your AI-pilled group chat come to life.

The feedback we heard was incredible. Founders shared that this is a huge season of transformation, and that making this space a reality helped everyone learn, move quickly, and accelerate by sharing what works, what breaks, and what people are learning today — even if it is probably outdated tomorrow.

One founder told us they walked away with “actionable insights that accelerated us right away” and new ideas that shaped their product vision.

The most impactful collective takeaway was reinforcing that everyone’s mental model for company building needs to be in a constant state of adaptation. What may be true today likely won’t be true in a month, so embrace the chaos as part of this transformation and find people and places that accelerate your thinking in real time.

On June 17, we’re bringing this to San Francisco.

This time, we’re calling it the AI Workflow Summit. And for the first time, we’re opening the room beyond the True community.

A Working Session, Not a Conference

Most AI events are still built around talking about what might happen next. Someone presents a workflow from a stage. A panel debates the future. A keynote explains where things are going.

But the people building with AI every day are already living inside that future.

They are not waiting for the next abstract conversation. They are rewriting workflows, team structures, communication norms, hiring criteria, management systems, and operating playbooks in real time.

They want to know what is working for the person at the other company right now.

• What skill?
• What new agent?
• How did you debug it?
• What does human-at-the-edge versus human-in-the-loop actually mean?
• Can I really launch a dark factory for my team?
• How are people managing collaborative agentic workflows?
• What about security, legacy code, and transforming culture in teams split between AI-pilled and AI-reluctant?

At the Accelerator, people did not come to perform. They came to compare what was actually working — and where they needed help.

We walked through live systems. Teams showed where things were breaking. People opened their laptops and shared the messy, practical details that rarely make it into the hype posts flooding the internet right now.

What Kept Surfacing

We’re not going to recap what individual founders shared. That belongs to them, and to the room. But across very different teams, a few themes kept coming up.

Code Is No Longer the Primary Artifact. The System That Generates It Is.

The founders pulling ahead are not simply writing more code. They are building the infrastructure around the code: prompts, validators, memory, orchestration, tests, review loops, and feedback systems that let output improve over time.

The artifact is not just the thing being produced. It is the mold that keeps producing better things.

Validation Is Becoming the Job

When models can generate enormous amounts of work, the question stops being, “Can we ship?” and becomes, “How do we know it’s right?”

Parallel Work Is the Future

One person can now send the same brief across multiple models, compare outputs, find the overlap, and investigate the disagreements. Those discrepancies are not noise — they’re signal to follow.

The Loop Is Moving From Human-in-the-Loop to Human-at-the-Edge

Agents run overnight. Work happens in batches. You wake up to review what was produced, not to start every task from scratch.

The interesting question becomes: Where does the human create the most leverage?

The Coding-Agent Pattern Applies to Every Function

Plan. Work. Review. Compound.

That loop is showing up far beyond engineering: marketing, research, sales, customer support, operations, community, and internal knowledge work.

The companies treating this as only a software engineering shift are missing the bigger point.

This is a company-design shift.

Management Skills Are Becoming More Valuable, Not Less

The people who can ask better questions, give clearer feedback, maintain taste, create trust, translate north-star vision, and structure work so humans and agents can both perform at a higher level are at a tremendous advantage.

Why the Room Matters

There is a reason this format is hard.

A lot of what builders are figuring out right now lives in the in-between: the small customizations, the half-working hacks, the brittle but useful workflows, and the moment someone says, “Wait, you can do that?” and the whole table leans in.

That knowledge does not travel well in a tweet.

It barely travels in a blog post.

It travels best in a room.

One founder texted us after the Accelerator:

“It’s a huge season of transformation and True is leading the freaking path on all of it. Thanks for making this space a reality.”

That is the bet behind the AI Accelerator and the upcoming AI Workflow Summit.

Being in the right community is part of the differentiator right now. Finding your people, learning quickly together, and staying a few weeks or months ahead matters.

The answer is in the room. We see it over and over.

June 17, San Francisco

The AI Workflow Summit is the next iteration: one day, demo-driven, hands-on, and built for people who want to work, not watch. If you’re a founder, builder, or operator trying to move faster with the tools shipping every day, we’d like you in the room.

Request an invite →

Come build with us.