lightcell energy πŸ”† (@lightcellenergy) header
lightcell energy πŸ”† (@lightcellenergy) profile picture

ultra power dense electric generator using light. portable dyson spheres DMs open

San Francisco
Joined August 2022
Photo by lightcell energy πŸ”† (@lightcellenergy)Photo by lightcell energy πŸ”† (@lightcellenergy)Photo by lightcell energy πŸ”† (@lightcellenergy)Photo by lightcell energy πŸ”† (@lightcellenergy)Photo by lightcell energy πŸ”† (@lightcellenergy)
lightcell CEO @daniellefong discusses the challenges that inventors must overcome with @benlandautaylor on @0xthestudents nitter.cf/0xthestudents/status/2…
Technology Doesn't Just Happen @DanielleFong, CEO of @lightcellenergy, talks about inventionβ€”how new technology depends on specific individuals, successfully not burning down the science shed while doing the early experiments, taking advantage of AI energy demand, and more. 0:00 β€” Intro 1:18 β€” Technology is not inevitable 5:22 β€” How Lightcell Energy became feasible 13:47 β€” The uses of theory, design, and experiment 20:26 β€” How better instruments and tooling fuels invention 27:07 β€” AI and the electrical grid 35:46 β€” How commercial demand interacts with inventor ideology 45:01 β€” Weaponize your weirdness 52:49 β€” What are your macros? 1:00:17 β€” Geothermal energy
2
10
779
lightcell energy πŸ”† retweeted
An underrated insight from cybernetics is that you can build something precise out of imprecise parts. Individual components can be inaccurate if you have a negative feedback loop, because the output depends mostly on how accurately you measure it, not the quality of the parts.
14
34
5
293
7,279
lightcell energy πŸ”† retweeted
πŸ™‚
after reviewing 10,000+ Thiel Fellowship applications, Michael and Danielle developed some pretty unusual ideas about what makes a great founder. edge control. hyperfluency. the toilet plunger. Danielle's "Friday Night Dyson sphere." none of these are a personality test. we're just looking for tricky bastards who always find a way.
2
20
1,194
lightcell energy πŸ”† retweeted
Technology Doesn't Just Happen @DanielleFong, CEO of @lightcellenergy, talks about inventionβ€”how new technology depends on specific individuals, successfully not burning down the science shed while doing the early experiments, taking advantage of AI energy demand, and more. 0:00 β€” Intro 1:18 β€” Technology is not inevitable 5:22 β€” How Lightcell Energy became feasible 13:47 β€” The uses of theory, design, and experiment 20:26 β€” How better instruments and tooling fuels invention 27:07 β€” AI and the electrical grid 35:46 β€” How commercial demand interacts with inventor ideology 45:01 β€” Weaponize your weirdness 52:49 β€” What are your macros? 1:00:17 β€” Geothermal energy
5
15
4
85
14,778
lightcell energy πŸ”† retweeted
pretty soon data centers as a net green energy asset gonna be the only way to get em done prospects
11
7
1
99
2,235
lightcell energy πŸ”† retweeted
it is honestly so eerie how the @lightcellenergy flame just floats above the nozzle. like a plasma ghost
3
5
1
33
3,243
lightcell energy πŸ”† retweeted
ran another advancement of our @lightcellenergy setup live, qualifying out new setup. first time we braved running the heat exchanger and oxygen enrichment (30-50%) at the same time! our new 3d prints worked, but slight assymetries in the torch passages lead to fuel concentration and some soot formation away from stochiometric next iteration, we'll really blast it clean before sintering
2
6
24
1,932
lightcell energy πŸ”† retweeted
I wrote an update on what we've done at Terraform this year, and what's next (roadmap.terraformindustries.…) terraformindustries.wordpres… It's a good day to make synthetic fuels.
27
31
16
436
25,541
lightcell energy πŸ”† retweeted
increasingly violent testing for our new @lightcellenergy setup, testing some new illuminants
5
5
1
77
4,365
lightcell energy πŸ”† retweeted
thank you, brilliant investors, for getting us this far, and beyond, on our epic journey. together we will shine brighter than the stars
i often describe @lightcellenergy as building portable dyson spheres. harnessing the power of spectrally pure, chemically fueled, man made stars. and it's remarkable that within just a few months of starting experiments, we started achieving brightnesses greater than the sun ever shines on earth. (note, in large part because it is 1/600,000th of our sky -- a limitation that our lightcells, our "dyson spheres" do not have, wrapping around our emitter.) within a few months, in a garden shed in Canada, we achieved over 3 suns, over 300,000 nits (cd/m^2). within a few more months, in our lab in San Francisco, Jon, Steve and I achieved almost 7 suns, 670,000 nits. albeit destructively to the apparatus! at temperatures that eroded our cutting torch, our salt supply, our quartz tubes, or cracked. our sapphire. but this was in 2023! to a large extent, the development over the past couple of years, from mid 2023, to 2024, and all this year, was a steady effort to invent and develop, in some cases for the first time, for any team on earth, a materials stack, and everything relevant to to, to make high temperature parts for (a) continuous operation, and (b) that could tolerate the extreme temperatures and chemical environment necessary to vaporize and condense salt, in a continuous cycle, with no moving parts. we basically continuously tried the stupidest thing that could possibly work. some things, like "use table salt" worked out amazingly well -- though they had knock on requirements, like now you have to reform and condense sodium chloride in heat exchangers other things, we tried all the stupidest things we could think of, until we ran out of stupid things to try. Then we had to start trying clever things. Like, the material that we chose, alumina, which can handle the molten NaCl, and the temperature, and will not further oxidize, since it is already an oxide, is super hard, as in high stiffness. It has finite thermal expansion. You must therefore relieve the thermal expansion via curvature. You can ONLY POSSIBLY survive thermal gradients and cycles through curvature. This will requires you pioneering a field. Roll up your sleeves. It has been an epic journey getting this far! Of course, it is yet unwritten if @lightcellenergy can even complete it, much less by the end of the year, but, i can certainly set a medium term goal: more than 1,000,000 nits (cd/m^2, or ~1kw/m^2). brighter than any firework. maybe achieved by New Years? that will be a hell of a firework, and a beautiful omen. 1000x brighter than HDR, 10x brighter than 1kW/m^2, 10 suns, the max brightness of a Sun, incident on the Earth's surface here's the twist: it would now not be cheating to use recuperation AND oxygen enrichment if you're burning green hydrogen from electrolysis, since you can set up to hold on the oxygen. it's for this reason, and many more that I'm excited to welcome my heroes, @TerraformIndies / @CJHandmer to @lightcellenergy's SAFE and Cap Table. We're building the Yin to @TerraformIndies' Yang of abundant hydrogen and synthetic fuels. We are excited to work on both the technology and the multi megascale to gigascale development sites -- energy campuses, power centers -- to build revolutionary, cost effective proving grounds for synthetic fuels and energy technology. Thank you so much. We are empowered and inspired by your belief and investment in us. We will make it count 🫑 Thank you again, as well, very much, to our many investors who we have not yet tagged, and will tag, on our THANK YOU FOR SUPPORTING US ON SAFES WHILE WE WERE JUST A DREAM post, soon to come. Hope to paint funding announcements in a new, golden light!
4
1
39
14,331
lightcell energy πŸ”† retweeted
This podcast inspired me to do something that's been on my bucket list for 15 years - add a geology layer to my mapping tool. Explore Cerro Gordo's tunnels and geology at maps.caseyhandmer.com.
In 2018, Brent Underwood did something nuts. He bought the old sliver mining town of Cerro Gordo for $1.4m. In so doing, he became the owner of a ghost town and many miles of mines. Brent has spent the last few years exploring the mines solo and trying to restore the town, which once played host to 4,000 people. Along the way, he’s become a YouTube star and a best-selling author. We went to Cerro Gordo to interview Brent and live at the mines for a couple of days. Here’s our full episode. The Core Memory pod is available at all fine pod retailers
6
5
109
6,216
lightcell energy πŸ”† retweeted
thanks so much to @benlandautaylor for this chance to talk about the development of new technology
Technology Doesn't Just Happen @DanielleFong, CEO of @lightcellenergy, talks about inventionβ€”how new technology depends on specific individuals, successfully not burning down the science shed while doing the early experiments, taking advantage of AI energy demand, and more. 0:00 β€” Intro 1:18 β€” Technology is not inevitable 5:22 β€” How Lightcell Energy became feasible 13:47 β€” The uses of theory, design, and experiment 20:26 β€” How better instruments and tooling fuels invention 27:07 β€” AI and the electrical grid 35:46 β€” How commercial demand interacts with inventor ideology 45:01 β€” Weaponize your weirdness 52:49 β€” What are your macros? 1:00:17 β€” Geothermal energy
9
14
1
103
9,850
lightcell energy πŸ”† retweeted
ramping up the new test rig!
12
6
1
113
8,364
lightcell energy πŸ”† retweeted
i can find the @nTopology design of @lightcellenergy photonic engines / ceramic compound nozzle / heat exchangers so engrossing that i do it all weekend! calls to me i guess. it kind of looks like a neoclassical column, don't you think? The design exercise was to do it with extrudable sections, which is potentially much more mass manufacturable than 3d printing. Of course, to test it out, we'll print it first. This compound nozzle core is an integrated air/hydrogen swirl nozzle and heat exchanger. It goes down the center of a quartz tube. The flow is ejected tangentially, making the flame a column that swirls around the core. The exhaust flow enters through the face, which uses the inversed swirl pattern at the front to tap the swirl energy and guide the exhaust flow in, after it diffuses. The internal blades and channels are composed of a single stretched sin wave, forming a spring-like compliant structure to accommodate thermal expansion, smoothly curved passages for nearly laminar swirling flow acceleration, lots of surface area for heat transfer, and manufacturability. A mirrored section on the opposite side swirls consistently to match a paired nozzle on the other side. An open gap allows alumina to expand while quartz stays fixed dimensions thermally.
11
10
63
5,688
hyperbowloid for khyber crystal experiments served us so well. nitter.cf/i/status/1981923064884… stay tuned for hyperbowloid 2
with @MaxLobovsky @formlabs CEO at the Formlabs Hardtech Meetup in SF showing off the @lightcellenergy hyperbowloid + sneak preview of its baptism by fire are going to use these fundamental structures in order to make a solid state internal combustion engine electric generator
2
7
932
lightcell energy πŸ”† retweeted
Please let us @lightcellenergy build your 2 GW+ of dyson sphere powerplant. We will use the sun when it's out and store our solar excess in hydrogen and oxygen, and then make solar cells illuminated with a salted vortex flame. we can store in tiered buffers like caches, L1 batteries, L2 Hydrogen, L3 synthetic fuels. We can do this @openai call me
Oracle just announced that it's seeking 2 GW of renewable energy to support its Stargate Project Jupiter data center. Here's some important context on this announcement: Oracle and OpenAI were initially planning to build 2.2 GW of natural gas generating capacityβ€”likely using relatively inefficient GE Vernova turbines. Those turbines could have emitted as much as 14 million tons of CO2 emissions per year. For context, 14 million tons of emissions would have wiped out 20 years of carbon emissions reductions in the state of New Mexico, where the project is based. For that, and other reasons, the project has been opposed by many state leaders and environmental groups. In March 2026, New Mexico’s Land Office blocked a gas pipeline that would have delivered natural gas to the project. Environmental groups then tried to convince FERC to slow down the permitting process in order to allow more time for public comment. In April, Oracle and OpenAI announced they were going to instead power the data center with 2.45 GW of Bloom fuel cells. As they touted in the press release, fuel cells emit much less NOx pollution than turbines. But they still run on natural gas and produce CO2 emissions. Even with the new plan the data center could be the largest source of emissions in the state. The project has continued to face opposition. That brings us to yesterday's announcement. Oracle is seeking 2 GW of renewable projects that can "deliver brand new renewable energy generation in New Mexico and support Project Jupiter’s commitment to achieve 100% carbon-free energy matching by 2031." Like all matching programs, this would be synthetic in the sense that 2 GW of renewables wouldn't directly power the data center. The Bloom fuel cells would be the primary power source, and Oracle would claim the environmental attributes of the new renewable projects located somewhere else in the state. At Cleanview we've been tracking Stargate Project Jupiter since it was announced. The image below is a satellite image taken about a month ago. We'll be continuing to track it through our data center tracker. Based on how this project has progressed so far, I expect many more updates to come over the next year.
3
9
1
42
3,106
lightcell energy πŸ”† retweeted
it's a big day for vortex generators. vortex generator generators, in particular @sama do you remember talking about fire tornadoes? similar. i should show you and olly some relevant things on that at @lightcellenergy we 3d printed these ceramic compound nozzles to be a steady rotation where the air and fuel are injected and merge smoothly, like a combined swirl nozzle and laminar flow element. it works weirdly well and makes a steady flow, and, channels and accelerates exhaust along the axis. it is very familiar to @openai's flow field, and i'm curious how similar!! and if such a flow field might be forced, and thus, exploited, in some actually physical invention. it doesn't seem like this basic idea of pulling the flow up the center, to do axial strengthing. and spiral nozzles, requires the continuum hypothesis to do vorticity amplification / concentration like this. so it should work in a device. similar high velocity axial vorticity amplification. in addition to high speed rotation stabilizing a density gradient like a flame column, it stablizes kinetically, in free air, and can be used for gas separation, particle extraction and energy pumping, like in a vortex tube (Ranque Hilsch vortex tubes at high and low temp ) and otherwise. not sure these are just my first thoughts seeing the paper your employee and my friend Abe Fetterman suggested double ended, which is the next thing we gotta try!
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
8
19
2
110
15,871
lightcell energy πŸ”† retweeted
holy crap, @ChatGPT Astra Ultra fucking nailed it. @nTopology implicits rendered in @Blender . Raytraced, with Alumina material. from @lightcellenergy hyperdome early prototype "midnight sun" by @DrSteveCrane
so, fable and sol already advanced our internal tooling so we had a better design thumbnail and search, rendering, info gathering and book keeping system. we had this posted just on our webdomain, locked behind @workos auth. now i'm having astra try to represent meshes and ultimately nTop files as implicits! hopefully it works well enough that it becomes a useful translation layer to bootstrap context vocabulary --- ok it's getting there very quickly
15
14
3
357
41,063
lightcell energy πŸ”† retweeted
> wasian semiconductor manufacturing company i can do that
company names in 2027: - dutch east chindian company - wasian semiconductor manufacturing company - the browser company of windia
1
2
54
3,054
lightcell energy πŸ”† retweeted
TIL we have achieved light flux density of approximately the surface of mercury. that is... curious. the solar cells for Maxeon 3 cells (Sunpower) were optimized for 7 suns as well. it is very suggestive, from the simulators. it is as if they are saying, listen. disassemble mercury. it is right there
2
3
50
3,432