FreshRSS

Normální zobrazení

Jsou dostupné nové články, klikněte pro obnovení stránky.
PředevčíremHlavní kanál
  • ✇Ars Technica - All content
  • Human muscle cells come back from space, look agedJacek Krywko
    Enlarge / Muscle atrophy is a known hazard of spending time on the International Space Station. (credit: NASA) Muscle-on-chip systems are three-dimensional human muscle cell bundles cultured on collagen scaffolds. A Stanford University research team sent some of these systems to the International Space Station to study the muscle atrophy commonly observed in astronauts. It turns out that space triggers processes in human muscles that eerily resemble something we know very wel
     

Human muscle cells come back from space, look aged

2. Srpen 2024 v 17:33
Image of two astronauts in an equipment filled chamber, standing near the suits they wear for extravehicular activities.

Enlarge / Muscle atrophy is a known hazard of spending time on the International Space Station. (credit: NASA)

Muscle-on-chip systems are three-dimensional human muscle cell bundles cultured on collagen scaffolds. A Stanford University research team sent some of these systems to the International Space Station to study the muscle atrophy commonly observed in astronauts.

It turns out that space triggers processes in human muscles that eerily resemble something we know very well: getting old. “We learned that microgravity mimics some of the qualities of accelerated aging,” said Ngan F. Huang, an associate professor at Stanford who led the study.

Space-borne bioconstructs

“This work originates from our lab’s expertise in regenerative medicine and tissue engineering. We received funding to do a tissue engineering experiment on the ISS, which really helped us embark on this journey, and became curious how microgravity affects human health,” said Huang. So her team got busy designing the research equipment needed to work onboard the space station. The first step was building the muscle-on-chip systems.

Read 11 remaining paragraphs | Comments

  • ✇Ars Technica - All content
  • DARPA’s planned nuclear rocket would use enough fuel to build a bombJacek Krywko
    Enlarge (credit: OLE-CNX) High-assay low-enriched uranium (HALEU) has been touted as the go-to fuel for powering next-gen nuclear reactors, which include the sodium-cooled TerraPower or the space-borne system powering Demonstration Rocket for Agile Cislunar Operations (DRACO). That’s because it was supposed to offer higher efficiency while keeping uranium enrichment “well below the threshold needed for weapons-grade material,” according to the US Department of Energy. This ju
     

DARPA’s planned nuclear rocket would use enough fuel to build a bomb

10. Červen 2024 v 20:56
A lump of rock, next to the periodic table entry for uranium, all against a black background.

Enlarge (credit: OLE-CNX)

High-assay low-enriched uranium (HALEU) has been touted as the go-to fuel for powering next-gen nuclear reactors, which include the sodium-cooled TerraPower or the space-borne system powering Demonstration Rocket for Agile Cislunar Operations (DRACO). That’s because it was supposed to offer higher efficiency while keeping uranium enrichment “well below the threshold needed for weapons-grade material,” according to the US Department of Energy.

This justified huge government investments in HALEU production in the US and UK, as well as relaxed security requirements for facilities using it as fuel. But now, a team of scientists has published an article in Science that argues that you can make a nuclear bomb using HALEU.

“I looked it up and DRACO space reactor will use around 300 kg of HALEU. This is marginal, but I would say you could make one a weapon with that much,” says Edwin Lyman, the director of Nuclear Power Safety at the Union of Concerned Scientists and co-author of the paper.

Read 15 remaining paragraphs | Comments

  • ✇Ars Technica - All content
  • Exploration-focused training lets robotics AI immediately handle new tasksJacek Krywko
    Enlarge (credit: boonchai wedmakawand) Reinforcement-learning algorithms in systems like ChatGPT or Google’s Gemini can work wonders, but they usually need hundreds of thousands of shots at a task before they get good at it. That’s why it’s always been hard to transfer this performance to robots. You can’t let a self-driving car crash 3,000 times just so it can learn crashing is bad. But now a team of researchers at Northwestern University may have found a way around it. “Tha
     

Exploration-focused training lets robotics AI immediately handle new tasks

10. Květen 2024 v 20:22
A woman performs maintenance on a robotic arm.

Enlarge (credit: boonchai wedmakawand)

Reinforcement-learning algorithms in systems like ChatGPT or Google’s Gemini can work wonders, but they usually need hundreds of thousands of shots at a task before they get good at it. That’s why it’s always been hard to transfer this performance to robots. You can’t let a self-driving car crash 3,000 times just so it can learn crashing is bad.

But now a team of researchers at Northwestern University may have found a way around it. “That is what we think is going to be transformative in the development of the embodied AI in the real world,” says Thomas Berrueta who led the development of the Maximum Diffusion Reinforcement Learning (MaxDiff RL), an algorithm tailored specifically for robots.

Introducing chaos

The problem with deploying most reinforcement-learning algorithms in robots starts with the built-in assumption that the data they learn from is independent and identically distributed. The independence, in this context, means the value of one variable does not depend on the value of another variable in the dataset—when you flip a coin two times, getting tails on the second attempt does not depend on the result of your first flip. Identical distribution means that the probability of seeing any specific outcome is the same. In the coin-flipping example, the probability of getting heads is the same as getting tails: 50 percent for each.

Read 16 remaining paragraphs | Comments

❌
❌