Energy, space, and the long-horizon research that does not fit a quarterly report.
The epidural device helping a paralysed man close his hand is real and carefully built. The national 'brain-chip' industry being announced around it is running years ahead of what the electrodes have shown.
The Nancy Grace Roman Space Telescope launches Sunday on a Falcon Heavy. Follow the money and it is the cheapest way NASA has found to build a great observatory — which is exactly why it almost didn't survive the budget.
Neural operators are a genuine advance in how machines learn continuous systems. Whether "five trillion data points in a single prompt" means anything yet depends on a test the company hasn't shown.
LandSpace became the third company anywhere to set an orbital-class first stage back down intact. It is also reportedly burning something like $240 million a year, and the entire case for the landing is a number it hasn't hit yet.
Two Falcon 9s, opposite coasts, half an hour apart. The number that matters isn't the clock on the doubleheader — it's what a launch costs once you can do it that often.
Fuse's dense plasma focus fired 1.27 trillion neutrons from a single pulse on a venture budget — a real efficiency record. It measures how hard the plasma pinched, not energy out, and the most honest thing the company does is sell the flash, not a power plant.
Space-domain awareness used to mean a handful of exquisite government satellites. Astranis is betting the whole business flips when the watcher costs a tenth as much and rolls off a broadband assembly line.
A risky maneuver engineers nicknamed the Big Bang found a year of science in the difference between two ways of running 49-year-old hardware. The reason it's worth doing is that the replacement costs infinity.
A 582-tonne coil colder than deep space, made without a single imported part, is a genuine engineering milestone. It is also a cage — and the field's hard problem was never the cage.
AST SpaceMobile can put broadband on an ordinary phone from orbit. The engineering is real. Whether the spreadsheet closes before the cash does is a separate, harder question.
All four big LHC experiments now report indications of quark–gluon plasma in collisions of oxygen and neon — nuclei far lighter than the lead they usually smash. The measurement is real. Whether a droplet of the primordial fluid formed is the careful part, and the physicists chose their words with more care than the headlines did.
A proposed FAA rule would let the agency waive 13 environmental laws for rocket launches. Read it as a spreadsheet: it lifts a cost off the launch companies' books and moves it onto the wildlife refuge next door.
Hitachi, Intel and AIST are betting that scalable quantum computing runs through existing silicon factories, not bespoke lab rigs. The manufacturing logic is sound. The thousand-qubit milestone sits a year past the money.
Starship Flight 13 is set to carry 20 Starlink V3 satellites — the first real payload on a test vehicle. They are too big for Falcon 9, and that is the whole business case.
The surgery restored no function on the day, and the device has eight electrodes against Neuralink's thousand. What China built first was the reimbursement plumbing — a year before there was anything to bill through it.
General Fusion rang the Nasdaq bell this week as the first publicly listed fusion company. Its machine has reached 8.4 million degrees — and needs roughly fourteen times that, plus two things it has never shown, before the word "power" means anything.
The Moon Base Program is a bet that cheap, repeatable deliveries beat one perfect mission. The math only closes if the landing rate climbs faster than the price.
Every kilogram of landing gear a rocket carries is a kilogram of payload it doesn't. The Long March 10B moved the landing gear onto a boat — and the budget for a Moon landing is the reason why.
City Labs' BOHR reached orbit this week carrying a tritium battery decades old in principle. What's actually new is a piece of paper — and the paper may be worth more than the payload.
Kate Adamala's SpudCell completed a real, chemically defined cell cycle — feeding, growing, dividing across five generations. What it hasn't completed yet is peer review, or a life of its own.
The Iridium deal isn't about rockets. It's about a subscription business, a defense contract, and a block of spectrum that no booster can put into orbit.
A new US facility will try to build the first full-scale fusion breeding blanket — the part that has to survive the neutrons, capture the heat, and make its own fuel. That it is only now being designed tells you how far "net energy" still is from a power plant.
Its new disc-shaped capsule can return 1,000 kilograms from space, many times what today's vehicles manage. The wager isn't the rocket. It's that bringing cargo home is finally getting cheap enough to build factories in orbit.
A startup's robot is about to grab a satellite that was never built to be grabbed. If it works, the real product isn't a rescued telescope — it's a price for keeping things in orbit alive.
Artemis III now flies to Earth orbit to rehearse two rival landers without touching the Moon. The astronauts are the headline. The story is a roughly $10 billion bet on competition — and the contender that just lost its only launch pad.
IBM let a language model search a space no human could and it found error-correction codes that beat the textbook. The distance between a better blueprint and a working machine is still measured in years.
The AI1 satellite is a real engineering answer to a real grid problem. Whether it ever deploys comes down to the oldest number in energy: cost per kilowatt-hour, delivered.
Washington just finalized a national plan to put fusion on the grid within a decade. The science behind it is real and the document is unusually honest — which is exactly why its most truthful sentences are the disclaimers.
The largest IPO in history prices Thursday and trades Friday at a $1.75 trillion valuation. Strip away Mars, and what you are actually buying is a dish on a roof — and a booster that lost an engine last week.
The prospectus asks for $1.75 trillion and points at a rocket. Read the segments and the rocket is the loss leader — the company is an internet provider that happens to fly its own delivery trucks.
Starship V3 made a fiery splash in the Indian Ocean and Elon Musk called it epic. The number that matters is dollars per kilogram to the lunar south pole — and that number is still doing the heavy lifting NASA's renderings won't.
Thea Energy’s oversubscribed Series B is real money for a real machine. The interesting question is what the money is actually buying — and on what timescale.
Commonwealth Fusion has applied to plug a 400-megawatt fusion plant into the grid. The reactor it depends on has not yet produced a single watt of net energy — and that is the whole story.
