HN user
24e6715b6543dfb02088ea00df72e2fd666224247c848776af21ef6039d44e32
Incredibly well executed Youtube project IMO. The DIY many-poled rotary encoders are particularly gorgeous cost engineering in a hobby project like this.
Fun old project but the technology has improved[0] since then.
I have a fun omega-3 anecdata point going right now. A friend of mine researches the stuff in mouse models and told me it's extremely beneficial, but you need to buy a fancy brand to avoid rancid oil or heavy metal contamination both of which ~null out the benefits. She recommended Sports Research.
I bought some and started taking it and my 1:1 bullet chess ELO jumped from 850 to ~1070 over the next couple weeks.
I play chess a bit like sushi ginger for the mind - purge working memory with a short intense task to context switch. I intentionally don't study openings or anything so I can use it as a benchmark for mental horsepower with a reasonably slow drift in the baseline from 'actually learning chess'.
My friend says this effect is way too big to actually attribute to the vitamins and it has to be placebo etc but I'm thoroughly enjoying the idea that omega-3 Nick would win 3/4 bullet matches against deficient Nick.
https://www.chess.com/member/nickparkerprint/stats/bullet?da...
It works if you just plainly describe what you're looking for, I write a new prompt for different images just like "re-render this as a matte untextured 3d model, remove all details except geometric form"
I've had surprising success with meshy.ai as part of a workflow to go from images my friends want to good 3D models. The workflow is
1. Have gpt5 or really any image model, midjourney retexture is also good, convert the original image to something closer to a matte rendered mesh, IE remove extraneous detail and any transparency / other confusing volumetric effects
2. Throw it in meshy.ai image to 3D mode, select the best one or maybe return to 1 with a different simplified image style if I don't like the results
3. Pull it into blender and make whatever mods I want in mesh editing mode, eg specific fits and sizing to assemble with other stuff, add some asymmetry to an almost-symmetric thing because the model has strong symmetry priors and turning them off in the UI doesn't realllyyy turn them off, or model on top of the AI'd mesh to get a cleaner one for further processing.
The meshes are fairly OK structure wise, clearly some sort of marching cubes or perhaps dual contouring approach on top of a NeRF-ish generator.
I'm an extremely fast mechanical CAD user and a mediocre blender artist, so getting an AI starting point is quite handy to block out the overall shape and let me just do edits. EG a friend wants to recreate a particular statue of a human, tweaking some T-posed generic human model into the right pose and proportions would have taken me "more hours than I'm willing to give him for this" ie I wouldn't have done it, but with this workflow it was 5 minutes of AI and then an hour of fussing in Blender to go from the solid model to the curvilinear wireframe style of the original statue.
Bad popularizing article, doesn't cover the actual conclusion:
The main points emerging from the combined simulation and experimental study on atmospheric entry of the paper plane are: • Orbit: The paper space plane de-orbits from LEO extremely quickly due to its very low ballistic coefficient. Atmospheric entry from a 400 km circular orbit occurs within a few days. • Attitude: In the free-molecular portion of atmospheric entry, above ∼120 km altitude, the paper space plane maintains a stable flow-pointing attitude. Small-amplitude oscillations occur in pitch and yaw. Although the coupled simulator is not designed for application at lower altitudes, the results suggest the onset of uncontrollable tumbling at ∼120 km altitude. • Heating: Based on the hypersonic wind tunnel test results and simulation, surface forces acting on the space plane during atmospheric entry are not expected to cause significant deformation. However, the paper space plane experiences severe aerodynamic heating in the order of 105 W/m2 (or 10 W/cm2 ) for several minutes. Accordingly, combustion or pyrolysis is expected during atmospheric entry
Also the GOAT of cable laying articles: Neal Stephenson doing gonzo journalism on the topic in the 90s
https://euripides.dk/setebos/frx/matrix/ai/books/stephenson_...
I found myself reading this more for the excellent notetaking than for the content.
I suspect the discussion was about as charged, meandering, and nitpicky as we all expect a PL debate among deeply opinionated geeks to be, and Jake Edge (who wrote this summary) is exceptionally good at removing all that and writing down substance.
RO desal splits the incoming ~3% salinity stream into two halves, one fresh and one ~6% salinity. This concentrated brine is fed to adjacent brine processing facilities (ideally in both countries) that exploit the region’s abundant solar and geothermal energy to extract potentially millions of tonnes of lithium, sodium, magnesium, chloride, and other metals found in sea water. The resulting depleted brine is piped back to the ocean where it is thoroughly mixed with sea water and discharged.
Casey's proposing we mine the brine for useful minerals. You're right he's glossing over details, but a citation addressing the economics of brine disposal with his proposed processing would add more to the discussion.
IIRC we've found some big exoplanets in wild orbits. They tend to form in a circular-ish orbit because that's how you get a big enough chunk of the dense bits of an accretion disk, but interactions with other planets/stars can put eg gas giants into wild orbits after formation.
EG https://astronomynow.com/2019/08/28/exoplanet-found-in-unusu...
Please also be careful about selling to incumbents without a strong guarantee (ie contractual penalties) your product will stay in market for many years.
I know smoke detectors and a couple other crappy home goods have been fixed by scrappy upstarts, sold to the losers for <100M, and then shuttered to avoid pivoting the existing business. Heartbreaking lost progress every time it happens.
Almost nobody in tech is interested in the degrowth solution to climate change, we’re all in on increased affluence with tech mitigations for the environmental harms.
Somebody’s gotta book all the electric / / synfuel powered flights ;)
Active vacuum insulation isn’t a thing because you need high vacuum to reduce thermal conductivity at all, because mean free path and therefore specific conductivity grows at the same rate density decreases until mean free path exceeds the characteristic length of your application.
Also, modern VIPs have decades of life if installed properly.
Anyone curious about psychedelic methods of action, particularly based on their personal experience, should read Surfing Uncertainty or at least the SSC review/summary of it.
The predictive processing interpretation of these drugs is that they change the balance between top-down predictive generative models from our "deep" brain and noisy bottom-up sensory percepts from the peripheral brain. When you get super-resolution like experiences of banal everyday textures, that's your peripheral brain failing to filter them out for lacking novelty, when you get psychedelia style visuals that's intermediate "layers" of abstraction running amok, and when you go off on extended narrative experiences fully detached from reality that's the deep generative models doing their own thing unmoored from the error signals your senses would otherwise generate.
It's an awesome book in many other respects, but IMO the best bit is how many plausible mechanistic explanations it offers for psychedelic experience.
https://slatestarcodex.com/2017/09/05/book-review-surfing-un...
Jane Street recruits silly hard from Cornell’s CS dept because part of our required curriculum is functional programming w OCaml. They definitely introduce a lot of math/cs kids to the idea that finance can be a meaningful technical challenge instead of just Dyson bros in spreadsheets.
Then again I think one of the founders or top execs is an alum, so it’s possible Cornell has that course in that language because of Jane Street
Water also has an almost uniquely large enthalpy of melting. EG paraffin freezing releases 1/3 as much heat as water
The tether article a week or two ago described them lending tremendous amounts of USD to exchanges and various crypto bank-like entities with Bitcoin as collateral
“ I also learned that Tether had lent billions of dollars more to other crypto companies, with Bitcoin as collateral. One of them is Celsius Network Ltd., a giant quasi-bank for cryptocurrency investors, its founder Alex Mashinsky told me. … Hoegner, Tether’s lawyer, [said] … that its secured loans are low-risk, because borrowers have to put up Bitcoin that’s worth more than what they borrow.”
https://www.bloomberg.com/opinion/articles/2021-10-07/matt-l...
Has anybody reverse engineered the interface for Apple's camera modules, particularly the front modules with LIDAR?
They can be had for $10 a pop on Alibaba now, and they're much nicer than similarly priced webcams let alone LIDARs.
Seems like a part that's ripe for a raspberry pi interface board.
In the original General Fusion plan: Pneumatic pistons which strike anvils in the chamber wall creating a shockwave which implodes the liquid wall with fusion-igniting pressures.
I think they may have moved away from that in favor of big pneumatic pistons pushing tiny piston heads directly into the liquid though. Mechanical advantage is the area ratio, which you can easily make quite large.
Scrolled pretty far and haven’t seen mention of this:
Water shortages will be entirely solved in the next decade by cheap renewables and desalination.
Clean water can be manufactured, and effectively free electricity will make it dirt cheap
This process has nothing to do with sintering - it's a new material for FDM that supports electroless nickel plating.
The hypothetical medical application is probably to seal a surface and/or apply functional coatings, eg various antimicrobial structures.
I want somebody to isolate these, re-engineer them to respond to a control input, and use them to implement nanowire network reservoir computing
https://www.sciencedaily.com/releases/2019/12/191226084403.h...
Casey Handmer argues [0] that The best solution is an “air mattress” structure where anchor cables are spaced throughout the structure to resist the pressure load.
The anchor cables can scale linearly with area just like the pressure does, while a perimeter wall anchor scales with radius and therefore sets a limit on enclosure size.
The cable anchor approach also lets you build arbitrarily high ceilings, like kilometers up hypothetically.
I really wish someone would build one of these on Earth. It would actually be easier because you’d just set the pressure high enough to resist wind loads. You could make an incredibly gigantic greenhouse in some otherwise inhospitable climate
[0] https://caseyhandmer.wordpress.com/2019/11/28/domes-are-very...
Hmm yeah that does sound like a permanent change in the second link doesn't it? Too bad, I was really rooting for their plan.
Can you point me to where they say the central cones are a change in their hypothetical production design?
I was under the impression it's just there for diagnostics in the sub-scale prototype.
Huh, I suppose you do! I had a naive short-circuit in my head that "neutron flux is bad because it destroys your your materials. If your material is molten lead anyway who cares?"
I also didn't know lead has such a low neutron absorption cross section.
For what it's worth lead with extra neutrons seems a lot less scary than eg uranium fission waste. Stable isotopes 206, 207, and 208 represent 98% by abundance, 208 has the lowest cross section, and 209 has a 3 hour half-life into Bismuth-209 which is nearly stable (2e19 year half-life). So it seems almost all of your neutron captures just make other stable lead isotopes or briefly-terrifying 209 that's totally safe after a couple days. You only get real scary stuff if the trace amounts of undecayed 209 manage a second capture.
Edit: I should add they want to mix lithium into the lead to absorb neutrons and regenerate fuel
Have you seen General Fusion?
They spin a large ball of molten lead using pumps and shoot hot plasma into the vortex that forms in the middle, then strike the walls of the lead chamber with carefully timed steam pistons to make an implosion pressure wave not entirely unlike Fat Man’s design, except reusable and for fusion.
The fusion heat and radiation gets absorbed by the lead and they run a heat exchanger on the pump loop for power generation. They also think they can breed fuel with some lithium in the lead.
I love their plan aesthetically but I’m not qualified to judge how feasible it is. It’s very satisfyingly physical though.
I want somebody to figure out a way to "backprop" / train neuromorphic nanowire circuits[0] and interface one with this retina to create a complete bionic visual system.
The main limitation of the retina appears to be interfacing the nanowires to a processor, and doing most of your processing in trained nanowires instead of digital circuitry maybe solves that.
You could make a sensor with all the benefits of human vision and similarly minuscule power consumption.
[0] https://www.sciencedaily.com/releases/2019/12/191226084403.h...
H/t the excellent Orbital Index newsletter.
This is the third ever NIAC Phase III grant, which is $2M over two years to transition the most promising advanced concepts to other parts of NASA.
The mission would fly a swarm of imaging satellites to 550 AU (11x Pluto's maximum orbital radius) using solar sails and ultimately image an exoplanet directly with a surface resolution of 25 km per pixel (!).
NIAC also funded a phase 2 grant for a new high performance solar sail which could reach the imaging point in just 8.5 years.