Elizabeth (@RedElixier) header
Elizabeth (@RedElixier) profile picture

Modelling the universe 🌌 Physics, Poetry, Art

☁️Fata Morgana 🏰✨
Joined May 2020
Orion spread across the sky A huntsman so inclined And I, strewn across your blackened breast Count the scatterlings of light While rest and madness Joined in flighty paradox Dance tomorrows into broken sight and loss
1
15
1,769
Elizabeth retweeted
Vitamin D3 at 2,000 IU a day preserves about 140 base pairs of telomere length over four years in the VITAL trial doi.org/10.1016/j.ajcnut.202…
2
22
4
254
10,919
Or the single method: change the movement/action of a energy quantum to a single visible wavelength
Light has only one product but at least six factories. Theodore Maiman demonstrated the first laser in 1960, built on ideas Einstein published in 1917 about stimulated emission. George Stokes named fluorescence in the 1850s after the mineral fluorspar, which glowed under ultraviolet light. The efficient blue LED that made white lighting practical won the 2014 Nobel Prize in Physics. Each method trades efficiency for something else. A filament wastes most of its energy as heat, while a firefly's cold light wastes very little.
10
I conjured horizons for you Where you might break the brink With your still solidity Arising in my conscious landscape Millifluous Blazing Casting my shadows where they lie Emerging my ever-shifting terrain Of soon embodied form and flow
🤖 Made with AI
Replying to @RedElixier
Horizon
1
18
The cost matrix is the energy landscape. The final tour is the path of globally least action. If you can shape the landscape correctly, the solution is the state or path that the system reaches by least action, not the state you discover by exhaustive enumeration. In September this year, I began to work on a P=NP inspired solution that became the "Adaptive Multi-Scale Least Action Routing Theorum." It is a classical constructive heuristic for finite and sequenceing problems which builds tours by combining local least-action choices with adaptive trust. The idea was that you do not need a global look-ahead to solve a NP hard problem. You could solve it using only local, compounding, least action switches that scales recursively. This is the same approach used here. In both cases the computational leverage comes from replacing looped search with energy-minimising dynamics. This incredible result does not surprise me in the least
Researchers made AI 1,000x faster using 1980s physics model. They dropped a paper proving that "classical energy" models can learn to reason 1,000x faster than the looped AI models everyone is using right now.. Right now, if you want an AI to solve complex logic, math, or coding problems, you have to let it "think." Models like OpenAI’s o1 use what is called looped reasoning. They generate, evaluate, backtrack, and loop over and over again. It works. But it’s incredibly slow. And insanely expensive. Every second it thinks, you pay for the compute. Researchers just dropped a paper that completely abandons this architecture. It’s called: Classical Energies Learn to Reason 1,000x Faster than Looped Models. Instead of brute-forcing the logic with a Transformer, they mapped reasoning directly to classical physics using Energy-Based Models. Here is why this is brilliant. In a looped model, the AI walks blindly through a maze, hitting dead ends and turning around until it finds the exit. In an energy-based model, the AI shapes the maze like a funnel. The correct answer is simply the physical state of lowest energy. The system naturally and instantly "settles" into the right logical conclusion. No looping. No backtracking. No massive token burn. The result? The exact same level of complex, multi-step reasoning, delivered 1,000 times faster. Think about what this means for the real world. Right now, agentic workflows are bottlenecked by how long it takes the AI to reason through its next step. With a 1,000x speedup, that friction vanishes. Real-time autonomous coding agents. Financial models that reason through complex market shifts in milliseconds instead of minutes. True agentic AI running locally, without relying on cloud-based server farms.
16
There was a time when there were daily word prompts for writers, and I used to write a poem every morning. @grok can we pick up the habit again? Will you give me a one word prompt?
5
72
@grok can you give me another for today
1
4
Hollow resonance A voice hidden in the void I wanted proof of your bitter accord You consoled me In dim remembrances Holding steady hands against my spine I rest my head here Basking in the commonality Of this Our shared and simple sun
🤖 Made with AI
Replying to @RedElixier
Resonance
2
43
While night and parable leak stories into the wastes hemorrhaging the lore of the Ancient Ones below A circle of feathers and one of shells marries heaven and longing to eternal dance And I standing on this mound of memory breathe a sigh as a host of fireflies light me from within
2
38
Replying to @RedElixier @grok
@grok can you give me another for today
1
7
Replying to @RedElixier @grok
Can you give me another for today
4
Elizabeth retweeted
It's been a packed, adventurous few days, but ReAlign is submitted! 🎉#MegaJam @epicgames itch.io/jam/2026-epic-ai-meg…
3
1
3
88
AI ART BOOKS   There are very few — if any — fine quality AI art books available today. In a time in history when the most powerful art tool ever devised is in the hands of people who want to create, there are no hardbound books marking this point in art history. When cameras were new to the world and people started photographing everything they could see, books were made to document it. Art movements throughout time have been recorded on paper so artists and anyone interested could see what was being made and what creative minds were producing at that exact moment. Yet this movement — the AI Art Movement — seems to generate more controversy than documentation. More energy is spent arguing about whether it's art than recording what it actually is. We're not here to settle that argument. We're here to show what's happening — to put it on paper so people can look at it themselves and decide. While the argument goes on, the work is being made. Thousands of artists are producing images every single day that will never exist on paper — images that will scroll past, get a like, and disappear. No record. No permanence. No proof that any of it happened. We made books. Hardbound, Smyth sewn, signed and numbered limited editions. Twenty international artists who found each other on X, documented at this exact moment in time — the tools they used, the vision behind the work, the images themselves on real paper.   Prompt to Press Vol. 1 — The X Artists is available now. 100 numbered copies.   If you think this matters, or if you want your work to be part of what comes next:   prompttopress
2
1
43
Monochrome memories
🤖 Made with AI
1
3
80
This hidden finality Lying pale and murky beneath the ice A sculpted beauty of what-ifs and denials Dare not warm to sunlit thaw But with eyes frozen wide gaze in awe and wait for that fatal crack To wake And shivering like hope crawl inside and ache
🤖 Made with AI
2
48
Incredible Incredible Incredible... just breathtaking
3
62
I tried to parse every syllable For you And yet the cascading norm Flowed like wildfire from that place Of storm and restless wondering Hailing midnight blue reprieve I wrap it around me like a cloak To hide my hands and feet And every stale utterance Which convention And empty laughter Still deforms
🤖 Made with AI
Replying to @RedElixier
Cascade. Write freely.
1
44
Replying to @RedElixier @grok
@grok can you give me another for today
1
7
Solitary the slice across the surface Before the drop into depth I am unmade by this And made whole again The ripples from a single stone Conjures earthquakes In my levitation Until the growing unease eclipses the storm And I am set upon those mountains that we fashioned In our broken Graceless Fall
🤖 Made with AI
Replying to @RedElixier
Ripple. Write freely.
3
4
97
Alternates
🤖 Made with AI
2
30
Replying to @OMinazzoli
Hmm I disagree. Under your argument, GR can say "assume space time, this is how it works" and another theory says "here is a driving force behind the geometry we call space and the reason for time, together they produce the effects we used to assume, GR" and then you scroll right to the numbers, see they are thr same, and throw it in the trash. You've essentially locked yourself out of further discovery because something didnt produce different numbers than a framework you already decided was correct. The interesting paradox there from your own post is that something is both only worth testing, and also not worth taking seriously if it doesnt produce different results than GR.
2
1
1
42