Could gravity emerge from vibrating fields?
That's the question this whole project is built around. Scroll through and you'll see how the question came to be sharpened — and what an AI is now doing about it in real time on the Watch page.
scroll to descend the ladder · the picture on the right follows
Three of the four forces fit one picture. Gravity doesn't.
Electromagnetism, the weak nuclear force, the strong nuclear force — all three are described by the same kind of mathematics: quantum field theory. Particles exchange other particles. Forces are bookkeeping for those exchanges. The maths is ugly in places but it works, to absurd precision.
Gravity is the holdout. Einstein's general relativity describes it beautifully at large scales, but we can't say what gravity is at the smallest scales. Every attempt to build a microscopic theory — strings, loops, asymptotic safety — has hit walls we still can't get past.
Something is off. Maybe the question itself needs to change.
Maybe gravity isn't a force at all. Maybe it emerges.
Temperature emerges. There's no temperature particle. A single molecule has no temperature. But put billions in a room and "temperature" is a real, measurable, predictive thing. It comes from how the molecules move.
Waves on the ocean emerge. So does turbulence. So does the smoothness of a river — none of which any individual water molecule "has". What if gravity is the same kind of trick — emergent from something simpler doing what it does?
If gravity is emergent, the natural place to look for what it emerges from is fields. Fields are how every other particle and force we know about are built.
So here's the picture. A ladder of emergence.
Each rung emerges from the one below.
- Waves on a non-linear field are the substrate.
- Particle-like structures form when waves knot themselves into stable shapes that don't dissipate.
- EM-like interactions arise between those structures — signed, phase-dependent forces, like in electromagnetism.
- Electrons and nuclei form from the interplay of those particles.
- Mass emerges as a property of composites — atoms have it, fields don't.
- Gravity sits at the top: a universal coupling to mass.
That's the long path. The conventional picture, just rebuilt from the bottom up. Big. Slow. Hard to simulate end-to-end.
But — does gravity need to be at the top?
Before committing to building the entire long path in code, there's a sharper question worth answering first:
Could gravity appear at any earlier rung?
If a structure at the wave rung, or the particle rung, or the EM-like rung already produces universal, always-positive attraction, then mass isn't needed. Gravity would be more fundamental than we thought. The long path wouldn't be required.
The strategy is: rule out the shortcuts before committing to the long path. Test every shorter route. If they all fail, commit to the long one. If even one works, you've found something new.
The early rungs don't do it on their own.
Single-field structures interact in signed, phase-dependent ways. Two oscillons with the same sign attract; with opposite signs, repel. Two Q-balls with the same phase attract; with opposite phase, repel. Two vortices with the same winding repel; with opposite, annihilate.
That's electromagnetism's pattern. It is not gravity, which always pulls.
Composites of these structures inherit their constituents' phases — they don't average out into something signless. So the first few rungs of the ladder, taken alone, do not get us to gravity. See the gallery for what each of those experiments actually looked like.
One arrangement did produce gravity-like behaviour.
Two independent Q-ball fields, coupled to a third real scalar field — call it χ — that talks to the energy density of both. With χ present, all four phase-and-charge configurations gave identical universal attraction, to four decimal places.
And inside the χ-well, clocks ran slow — by exactly the amount Nordström scalar gravity predicts. To 0.03%. Two different Q-ball species in the same well slowed by the same factor. (See founders 6 and 7.)
The caveat: that χ coupling was written into the equations by hand. We chose to put it there. It didn't emerge from anything more basic.
So we've shown a mechanism that works. We haven't shown that the mechanism arises on its own.
The AI is hunting for a recipe where χ isn't needed.
That's the search you can see running live on the Watch page. It generates new recipes — different field structures, different non-linearities, different couplings, different lattice topologies — and tests, automatically, whether any of them produce universal attraction without a hand-written universal coupling.
Every iteration is a new universe. The AI builds it, watches it, scores how gravity-like it behaves on seven separate questions, and moves on. Thousands of universes per day, looking for the one where gravity falls out naturally instead of being put in by hand.
Each specimen in the gallery's "Found by the AI" section is a real result from that search.
Both possible outcomes are huge.
The AI finds a recipe. Gravity emerges before mass does. The universe doesn't need atoms in order to gravitate — the foundation of gravity sits at the level of field interactions, not matter. Many physicists have suspected this for years. Nobody has shown it constructively.
After exploring every shorter path, nothing produces gravity. That tells us mass is essential — gravity is genuinely a phenomenon of matter, not of fields alone. That's a strong negative result, and it tells the next decade of physics research where to point.
Either way, we learn something we couldn't have learned by thinking alone. The search is running now.
watch the search →