Advanced Concepts Team

Works for mercury as well.

A flat metalens based on the control of the diffraction pattern of individual nanoantennas can achieve NA~1 bending light at angles as high as 82º

Neural Networks to analyse sleeplessness.

This has attracted a lot of attention in the DL community lately. Article: https://lilianweng.github.io/lil-log/2017/09/28/anatomize-deep-learning-with-information-theory.html Paper: https://arxiv.org/pdf/1703.00810.pdf

Graph coloring with an infinite number of vertices requires at least 5 colors (compared to the 4-color-theorem).

According to this site, there is not much coming. However, I like the idea.

The race for the AI chips intensifies.

A quick checklist to detect what sort of papers about blockchain are bullshit.

The B report shares the same conclusions and angle ...

Blockchains, now with more buzzwords!

Time travel! Finally. Can we also use it for antigravity?

A white creature with regeneration.

"Leading scientists have drawn up plans for a vast multinational European institute devoted to world-class artificial intelligence (AI) research in a desperate bid to nurture and retain top talent in Europe. The new institute would be set up for similar reasons as Cern, the particle physics lab near Geneva, which was created after the second world war to rebuild European physics and reverse the brain drain of the brightest and best scientists to the US."

"A paper posted online in January takes theoretical computer scientists halfway toward proving one of the biggest conjectures in their field. The new study, when combined with three other recent papers, offers the first tangible progress toward proving the Unique Games Conjecture since it was proposed in 2002 by Subhash Khot, a computer scientist now at New York University. Over the past decade and a half, the conjecture - which asks whether you can efficiently color networks in a certain way - has inspired discoveries in topics as diverse as the geometry of foams and the stability of election systems. And if the conjecture can be proved, its implications will reach far beyond network-coloring: It will establish what is the best algorithm for every problem in which you're trying to satisfy as many as possible of a set of constraints - the rules in a sudoku puzzle, or the seating preferences of a collection of wedding guests, for instance."

Researchers have used machine learning to predict the chaotic evolution of a model flame front.

Two ways of measuring the universe's expansion rate yield two conflicting answers. Many point to the possibility of new physics at work, but a new analysis argues that unseen errors could be to blame. See also this work based on GAIA data that, on the other hand, reinforces the discrepancy: Milky Way Cepheid Standards for Measuring Cosmic Distances and Application to Gaia DR2: Implications for the Hubble Constant https://arxiv.org/abs/1804.10655

A NASA study which is concerned with the question whether we could detect lost industrial civilizations on earth by analyzing the climate fingerprints.

Took 2000 GPUs over 2 weeks to train :)

Norwegian scientist Kristian Morten Olesen has patented a process to mix nano-particles of clay with water and bind them to sand particles to condition desert soil - he has been working on Liquid Nanoclay (LNC) since 2005.

The news here that some patent has been filed. If it actually works the guy is an ass for wanting to patent something that is solving world hunger. If it does not, then the guy is still an ass. So the news is that this guy is an ass :)

Agreed :) But, if it does work I doubt the guy will be able to enforce the patent in most of the countries where this will be helpful.