Let’s face it – you’re stuck inside. Why not read some hopeful news about what science can accomplish. This week we’ve got evidence of an ancient social network, hope for heart transplants, revelations about antimatter, an invisibility sweatshirt, and more.
A better competitor
If you’re stuck at home due to a lockdown order, you may be turning to technology to keep up your fitness routine. If you do, we have one piece of advice: don’t pump up your avatar.
Researchers at the University of Bath have discovered that people perform better in virtual reality exercise if they don’t create an aspirationally buff avatar of themselves on screen.
In the study, participants wore VR headsets and played a bike racing game. They competed against a generic avatar, an idealized avatar, and one that depicted them more accurately. The subjects performed better and found the game most motivating when they competed against the realistic version of themselves. When it came to the idealized avatar, racers had fun, but power output measurements showed that they didn’t try as hard.
Dr. Christof Lutteroth thinks the feedforward mechanism can explain the study’s results. This concept describes the performance gains and motivation a person sees when competing with a model that is easier to identify with. You believe you have a better chance of beating the real you, than beating the buff version of yourself. So if you want to meet your fitness goals in the coming weeks, stick with a realistic avatar.
How to be invisible
If you do go outside, you may be tracked. Video cameras are recording in stores, schools, and on city streets. Millions of households are equipped with devices like the Ring doorbell. Is there any chance we can escape this constant monitoring of daily life?
Some researchers, like the University of Maryland’s Tom Goldstein, say yes. They employ adversarial attacks to confuse the AI behind the cameras. One method Goldstein has developed is an “invisibility cloak” that renders the wearer undetectable to surveillance technology.
In this case, the cloak is a garish sweatshirt printed with a specially-designed pattern. They began with an image of random static, then added an image of people, and covered part of that with the static again.
After showing the image to a neural network, an algorithm updated the pattern to make it more effective. After many iterations on hundreds of thousands of images, the network could no longer recognize people when the adversarial pattern was present in the image.
The researchers themselves are not sure why the pattern works, only that it exploits weaknesses in the AI systems. Neural networks are only as strong as the data they are trained with, and it’s impossible to show them all possible images.
Deep sea connections
Half a billion years ago, organisms called rangeomorphs dominated the Earth’s oceans. Large and complex, they were our planet’s first non-microscopic lifeforms. Scientists recently studied 40 different fossil sites in Newfoundland, Canada where whole populations are available for observation. The fossils were well preserved because rangeomorphs could not move.
The researchers, led by Dr. Frankie Dunn of the Oxford University Museum of Natural History, discovered that a network of thread-like filaments connected rangeomorphs in an undersea “social” network. The filaments may have been used as a form of clonal reproduction. They may also have given the rangeomorphs stability against strong ocean currents and provided a means of sharing nutrients.
Dr. Alex Liu from Cambridge's Department of Earth Sciences, says that the organisms seemed to colonize the sea floor quickly. The filaments may explain how they were able to do that. Liu adds that scientists may need to reassess previous findings on how the creatures competed for resources on the ocean floor.
Partnering with pigs
About 98% of people in the US who need a heart transplant are not eligible to receive one. In an effort to increase the donor pool, researchers at the University of Minnesota Medical School are working to grow human-derived blood vessels in a pig.
Humans and pigs share similar physiological traits, and in the past scientists have successfully studied pigs to discover treatments for human conditions. For example, doctors treated diabetic people with pig insulin before human insulin was developed.
Making human blood vessels in pigs could allow scientists to make organs that are less likely to be rejected by the recipient. The blood vessels are created by injecting human-induced pluripotent stem cells, scraped from the patient’s own skin, into a pig embryo. Those future piglets would have blood vessels that match the patient exactly, eliminating the need for immunosuppression drugs.
The study’s co-author Mary Garry, says the proof of concept is solid; the next step will be to advance the research into the later gestational phase.
Why we're here
Matter-antimatter asymmetry is one of the great mysteries of physics. This phenomenon is also known as baryon asymmetry, and neither the standard model of particle physics nor the theory of general relativity can resolve the conundrum.
The standard model explains three of the fundamental forces in the universe: electromagnetism, the weak force, and the strong force. But it has less to say about the existence of dark matter, or a certain observed property of neutrons. So, in 1977, physicists proposed a particle called the axion to solve the neutron problem.
Now, researchers are suggesting that the hypothetical axion may also explain why the universe contains much more matter than antimatter. Raymond Co of the University of Michigan and his colleague Keisuke Harigaya at the Insitute for Advanced Study are rethinking an earlier idea that the axion field remained static just after the Big Bang. They suggest that the axion field had more interesting dynamics in the universe’s early stage.
The researchers think that when strong force and weak force interactions caused the axion to rotate, more matter than antimatter was created. Matter and antimatter came together and did not annihilate completely. Instead, one in 10 billion parts of matter was left to form the world we know today. Co says this axion framework will be tested experimentally in the near future.