THE MULTIVERSE

RESEARCH PROGRAM

SUMMER 2025 APPLIED AI SYSTEMS DESIGN

A three-month structured research initiative built to foster independent, exploratory work in applied AI, systems design, and speculative technology. This summer, our focus is on building.

Program Mission

The Multiverse Research Program is a three-month structured research initiative built to foster independent, exploratory work in applied AI, systems design, and speculative technology.

Participants are expected to produce real, working prototypes grounded in concrete research questions, while maintaining a collaborative ethos and commitment to learning in public.

This summer, our focus is on building. Writing is good. Thinking is good. But building systems—ones that push at the edges of what's possible—is better.

Program Timeline & Milestones

Phase 0: Soft Start (Now–June 1)

Goal: Get your ideas in shape and feel out your direction.

  • Begin drafting a research plan.
  • Submit early concepts: project title, short description, and a longer project narrative.
  • Optional: Start forming a research question or hypothesis.
  • Use this time to vibe code, experiment freely, and let ideas emerge.
  • Expect minimal support during this time—we're winding down Solarpunk, and encouraging you to build independence.

Phase 1: Research Commitment (June 1–June 15)

Goal: Lock in your direction.

  • By June 1, submit:
    • Project title
    • One-paragraph description
    • Long description or narrative
    • Research question (stated as a falsifiable hypothesis)
    • Initial experiment or build proposal
  • Between June 1 and June 15, you can change your plan—but you must finalize it by mid-June.

Phase 2: Build & Prototype (June 15–July 31)

Goal: Create a full research artifact.

  • Build a functioning system that:
    • Implements your hypothesis or research claim.
    • May require more compute, human labor, or tokens than you currently have.
  • Systems should be end-to-end complete, even if not scalable or optimized.
  • Write as you go—but write clearly. We will ask for:
    • High-level summaries.
    • Proposal-style writeups.
    • Diagrams, README-style explanations, and approachable documentation.
  • Expect to vibe code, use available models creatively, and push boundaries.

Phase 3: Publish & Deploy (August)

Goal: Translate research into impact.

  • Document and publish your project.
  • This could include: blog posts, GitHub repos, whitepapers, tutorials, recorded demos, etc.
  • Work with us to identify resource constraints (tokens, infra, etc.) and solve them.
  • Your final artifact should be a deployed prototype that others can understand, interact with, or build on.
  • Our goal: to help you finish strong and publicly share what you've created.

Proposal Format

You'll be expected to maintain a project record throughout the summer. This includes:

  1. Project Title
  2. Short Description (2–3 sentences)
  3. Long Description (~1 page of narrative or bullet points)
  4. Research Question (stated as a hypothesis)
  5. Experiment Proposal
    • What will you build?
    • What will you measure?
    • How will this test your hypothesis?

These materials should evolve over time. We will check in on them at three points: June 1, June 15, and August 1.

How We Work

Autonomy with Accountability

You're in charge of your timeline. We're here to help you succeed, not micromanage.

Build > Talk

We value concrete artifacts over speculative writing. Let your prototype speak.

Vibe Coding is Valid

Use intuition, creative coding, weird prototypes, and high-variance approaches.

Digestible Communication

We avoid dense academic speak. Clarity and accessibility matter.

Help Each Other

If you hit a wall, someone else may already have a ladder. Share early, share often.

Support

  • We'll provide weekly check-ins, lightly structured.
  • Guides and walkthroughs (e.g. "how to define your research question with Claude") will be provided.
  • Access to tokens, GPUs, or compute will be available for promising projects—we'll prioritize based on need and readiness.

Outcomes

By the end of this program, you will have:

  • A completed and documented research prototype.
  • A published artifact others can use, replicate, or build on.
  • A project writeup that could become a conference proposal, blog post, or grant submission.
  • Experience participating in a community of independent researchers pushing boundaries together.

If you're ready to experiment, build, and ask weird, specific questions—the Multiverse Research Program is ready for you.

Submit your project concept by June 1. Let's get to work.