# 🧪 PSI (Physical Superintelligence) — Research Package for Danielle
## AI + Physics: The Frontier Her Autonomous Submarine Work Points Toward
*Compiled March 31, 2026*

---

## What Is PSI?

**Physical Superintelligence PBC (PSI)** is a startup building the world's first "vertically integrated factory for physical superintelligence" — AI that doesn't just write text, but discovers new physics.

- **Founded by:** Dr. Alexander Wissner-Gross — Harvard/MIT physicist, computer scientist
- **Team roots:** Google, OpenAI, Harvard, Meta, MIT, Stanford, Oxford, Johns Hopkins, Cambridge, Institute for Advanced Study, Perimeter Institute
- **Locations:** Boston + San Francisco (expanding)
- **Website:** [psi.inc](https://www.psi.inc)
- **Mission:** "Discover and commercialize transformative physics breakthroughs at scale with artificial superintelligence — safely, verifiably, and for broad public benefit"
- **Motto:** *ignis scientiae, illumina mundum* ("the fire of knowledge, illuminate the world")

### Their Product: "Get Physics Done" (GPD)
An open-source AI copilot for physics research — the first agentic AI physicist.

- **GitHub:** [github.com/psi-oss/get-physics-done](https://github.com/psi-oss/get-physics-done)
- **What it does:** Takes a physics research question → scopes the problem → plans the work → derives results → verifies them → packages output
- **Installs into:** Claude Code, Codex, Gemini CLI, OpenCode
- **Install:** `npx get-physics-done`
- **Use cases:** Long-horizon physics research, numerical studies, manuscript writing/review, complex analytical work
- **Version:** 1.1.0 (released ~March 2026)
- **License:** Open source

### How It Works (Technical)
1. `new-project` — Start a fresh research workflow
2. `map-research` — Map existing work (formalism, computations, validation status, open questions)
3. `discuss-phase` → `plan-phase` → `execute-phase` → `verify-work` — Structured research pipeline
4. Multi-step verification against physical constraints — it checks its own work against nature's laws

---

## Why This Matters for Danielle

Danielle works at an **autonomous submarine startup** in Austin. Her engineering work sits at the intersection of:

1. **Physics** — Fluid dynamics, acoustics, navigation in 3D underwater environments
2. **AI/Autonomy** — Machine learning for decision-making without human input
3. **Robotics** — Physical systems operating in extreme environments

PSI represents the cutting edge of **physics-informed AI** — AI systems that understand physical laws, not just patterns in data. This is directly relevant to:

- **Autonomous underwater navigation** — physics models for currents, pressure, acoustics
- **Simulation engineering** — digital twins of submarine operations
- **Sensor fusion** — combining physics priors with sensor data for better state estimation

### Specific Opportunity
PSI is **hiring** (careers page: [jobs.ashbyhq.com/psi](https://jobs.ashbyhq.com/psi)). With her engineering degree from USC and autonomous submarine experience, Danielle would be an incredibly strong candidate if she's interested in the pure research side.

---

## The "Moonshot" Connection — Peter Diamandis / A360

Joe asked about the **A360 video** referenced on the Moonshot YouTube channel. Here's what I found:

### Moonshots with Peter Diamandis (YouTube/Podcast)
Peter Diamandis is the founder of XPRIZE, Singularity University, ZeroG, **and A360** (his executive community/mastermind).

**Dr. Alex Wissner-Gross appears on multiple episodes:**

1. **EP #208** — "The AI Wealth Gap: Why 40x Deflation Changes Everything"
   - With Dave Blundin (Link Ventures) and Salim Ismail (OpenExO)
   - Discusses $33 trillion deflationary collapse from AI
   - Wissner-Gross's thesis: AI will cause 40x cost deflation in many industries

2. **EP #209** — "What Everyone Missed About Gemini 3"
   - Same panel (Salim, Dave, Alex)
   - YouTube: [youtube.com/watch?v=r8RGYw1n-5k](https://www.youtube.com/watch?v=r8RGYw1n-5k) (Nov 20, 2025)

3. **"The Most Consequential Conversation I've Had in Two Decades"**
   - YouTube: [youtube.com/watch?v=TGwv9GZOFpY](https://www.youtube.com/watch?v=TGwv9GZOFpY) (Feb 19, 2026)
   - Deep dive on Wissner-Gross, PSI, and physics AI
   - **This is likely the "A360 video" Joe referenced**

### A360 Is Not a Separate Video — It's Diamandis's Community
"A360" is Peter Diamandis's paid executive mastermind/community for exponential entrepreneurs. The videos Joe saw are likely from the public Moonshots YouTube channel, where A360 content occasionally gets shared.

---

## What to Send Danielle

**Suggested message from Joe:**

> "Hey Danielle — check out this company PSI (psi.inc). They're building AI that actually understands physics — not just language, but the real laws of nature. Founded by an MIT/Harvard physicist. Their tool 'Get Physics Done' is open source on GitHub. Thought of you because of the autonomous submarine work — physics-informed AI is the next frontier for autonomous systems. They're hiring too if you ever want to go full research mode. Love, Dad"

**Links to include:**
1. PSI website: https://www.psi.inc
2. GitHub (Get Physics Done): https://github.com/psi-oss/get-physics-done
3. Moonshot video (the big one): https://www.youtube.com/watch?v=TGwv9GZOFpY
4. PSI careers: https://jobs.ashbyhq.com/psi

---

## Investor Angle (For Joe)

PSI is a **PBC** (Public Benefit Corporation) — private, not investable directly. But the physics-AI space is worth monitoring from an investment perspective:

- **NVIDIA** benefits massively (compute for physics simulation)
- **Synopsys** — engineering simulation AI
- **PhysicsX** (private, UK) — collaborating with Siemens + NVIDIA on physics AI
- The broader thesis: AI that understands physics = autonomous vehicles, drug discovery, materials science, energy breakthroughs

This sector validates Joe's **20-Year Innovation Holdings** bucket thesis. The companies building physics-aware AI will compound value for decades.

---

*Package ready to send to Danielle whenever Joe gives the green light.* 🦞