At 2 mins 26 seconds when I say “Peter” I meant “Paul”. Sorry!
Dark energy has got something to do with quantum vacuum fluctuation, whoa, physics. You have probably heard something like that. Alas, that isn’t quite right. In this video I clear up the confusion. Vacuum energy is much easier to understand than you might have been told. And it doesn’t fluctuate.
0:00 Intro. 0:27 Vacuum Energy according to Scientific American. 2:07 Vacuum Energy according to Sabine. 7:11 The Gas Analogy for Dark Energy. 10:04 Sponsor Message.
Using the European Southern Observatory’s Very Large Telescope (ESO
Created in 1962, the European Southern Observatory (ESO), is a 16-nation intergovernmental research organization for ground-based astronomy. Its formal name is the European Organisation for Astronomical Research in the Southern Hemisphere.
Astronomers may have seen the light from two black holes smashing into one another for the first time ever.
Black holes are completely dark and therefore invisible to light-detecting telescopes. So far, the only way astronomers have been able to “observe” black holes colliding is by detecting the resulting gravitational waves.
Deep Learning systems can achieve remarkable, even super-human performance through supervised learning on large, labeled datasets. However, there are two problems: First, collecting ever more labeled data is expensive in both time and money. Second, these deep neural networks will be high performers on their task, but cannot easily generalize to other, related tasks, or they need large amounts of data to do so. In this blog post, Yann LeCun and Ishan Misra of Facebook AI Research (FAIR) describe the current state of Self-Supervised Learning (SSL) and argue that it is the next step in the development of AI that uses fewer labels and can transfer knowledge faster than current systems. They suggest as a promising direction to build non-contrastive latent-variable predictive models, like VAEs, but ones that also provide high-quality latent representations for downstream tasks.
OUTLINE: 0:00 — Intro & Overview. 1:15 — Supervised Learning, Self-Supervised Learning, and Common Sense. 7:35 — Predicting Hidden Parts from Observed Parts. 17:50 — Self-Supervised Learning for Language vs Vision. 26:50 — Energy-Based Models. 30:15 — Joint-Embedding Models. 35:45 — Contrastive Methods. 43:45 — Latent-Variable Predictive Models and GANs. 55:00 — Summary & Conclusion.
ERRATA: - The difference between loss and energy: Energy is for inference, loss is for training. - The R(z) term is a regularizer that restricts the capacity of the latent variable. I think I said both of those things, but never together. - The way I explain why BERT is contrastive is wrong. I haven’t figured out why just yet, though smile
Scientists have been able to trap antimatter particles using a combination of electric and magnetic fields. Antiprotons have been stored for over a year, while antimatter electrons have been stored for shorter periods of time, due to their lower mass. In 2011, researchers at CERN announced that they had stored antihydrogen for over 1,000 seconds.
While scientists have been able to store and manipulate small quantities of antimatter, they have not been able to answer why antimatter is so rare in the universe. According to Einstein’s famous equation E = mc2, energy should convert into matter and antimatter in equal quantities. And, immediately after the Big Bang, there was a lot of energy. Accordingly, we should see as much antimatter as matter in our universe, and yet we don’t. This is a pressing unsolved mystery of modern physics.
According to Einstein’s equations, as well as other modern theories of antimatter, antimatter should be exactly the same as ordinary matter, with only the electric charges reversed. Thus, antimatter hydrogen should emit light just like ordinary hydrogen does, and with exactly the same wavelengths. In fact, an experiment showing exactly this behavior was reported in early 2020. This was a triumph for current theories, but meant no explanation for the universe’s preference of matter was found.
A team of researchers from TU Delft managed to design one of the world’s most precise microchip sensors. The device can function at room temperature—a ‘holy grail’ for quantum technologies and sensing. Combining nanotechnology and machine learning inspired by nature’s spiderwebs, they were able to make a nanomechanical sensor vibrate in extreme isolation from everyday noise. This breakthrough, published in the Advanced Materials Rising Stars Issue, has implications for the study of gravity and dark matter as well as the fields of quantum internet, navigation and sensing.
One of the biggest challenges for studying vibrating objects at the smallest scale, like those used in sensors or quantum hardware, is how to keep ambient thermal noise from interacting with their fragile states. Quantum hardware for example is usually kept at near absolute zero (−273.15°C) temperatures, and refrigerators cost half a million euros apiece. Researchers from TU Delft created a web-shaped microchip sensor that resonates extremely well in isolation from room temperature noise. Among other applications, their discovery will make building quantum devices much more affordable.
A dead star is spinning so rapidly, it officially has the fastest known spin rate of any star of its kind.
It’s a white dwarf star, named LAMOST J024048.51+195226.9 (J0240+1952 for short) and located 2,015 light-years away, and it has an insane rotation rate of just 25 seconds. That pips the previous record holder by a significant margin – CTCV J2056-3014, with a spin rate of 29 seconds.
It also bears a close similarity to another fast white dwarf, AE Aquarii, which has a spin rate of 33 seconds.
When we talk about the distance to an object in the expanding Universe, we’re always taking a cosmic snapshot — a sort of “God’s eye view” — of how things are at this particular instant in time: when the light from these distant objects arrives. We know that we’re seeing these objects as they were in the distant past, not as they are today — some 13.8 billion years after the Big Bang — but rather as they were when they emitted the light that arrives today.
But when we talk about, “how far away is this object,” we’re not asking how far away it was from us when it emitted the light we’re now seeing, and we aren’t asking how long the light has been in transit. Instead, we’re asking how far away the object, if we could somehow “freeze” the expansion of the Universe right now, is located from us at this very instant. The farthest observed galaxy GN-z11, emitted its now-arriving light 13.4 billion years ago, and is located some 32 billion light-years away. If we could see all the way back to the instant of the Big Bang, we’d be seeing 46.1 billion light-years away, and if we wanted to know the most distant object whose light hasn’t yet reached us, but will someday, that’s presently a distance of ~61 billion light-years away: the future visibility limit.
New ways to measure the top supercomputers’ smarts in the AI field include searching for dark energy, predicting hurricanes, and finding new materials for energy storage.
