Unfiltered image taken by Emlyn with his home made microscope (made with the lens of a laser pointer and an Android phone).

Originally shared by Emlyn O’Regan

Ok, this is something special now.

Originally shared by John Poteet

If this actually turns out to be true solar power will be the cheapest power source in the world. What a christmas gift. 

“A simple economic analysis shows that the proposed battery-to-solar-cell procedure could have a substantial impact. Assuming that the perovskite thin film is just half a micrometer thick, the researchers calculate that a single lead-acid car battery could supply enough lead for the fabrication of more than 700 square meters of perovskite solar cells. If the cells achieve 15 percent efficiency (a conservative assumption today), those solar cells would together provide enough electricity to power about 14 households in Cambridge, Massachusetts, or about 30 households in sunny Las Vegas, Nevada. Powering the whole United States would take about 12.2 million recycled car batteries, fabricated into 8,634 square kilometers of perovskite solar panels operating under conditions similar to those in Nevada.”

http://news.mit.edu/2015/solar-energy-discarded-car-batteries-1222

see the video here: 

https://www.youtube.com/watch?v=LP9HmTrUms0

h/t Keith Wilson 

The Internet of Shit

http://motherboard.vice.com/read/when-the-internet-of-things-starts-to-feel-like-the-internet-of-shit

A 10^8 improvement over classical computing is pretty good, you know. The idea behind simulated annealing is to search through a noisy solution space finding a good solution—though not necessarily the best solution—as quickly as possible. If you see the solution space as a (massively multi dimensional) landscape then the better solutions are like mountains in that landscape.

Imagine searching a landscape of many hills for the highest mountain, also imagine that you have to search this landscape blind and by foot. There are several methods you could use. The simplest is: always go up. That will get you up the nearest hill but it’s not likely to get you up one of the highest ones. For that you need to find intelligent ways to go down as well as up. Annealing is a probabilistic way improve the search: at the beginning go up and down often but gradually increase the tendency of going up more often that down. This enables a rapid exploration of the solution space while ratcheting up the preference for hill climbing and staying with good solutions. https://en.m.wikipedia.org/wiki/Simulated_annealing

Quantum data annealing enables exploring countless solutions simultaneously by exploiting the probabilistic properties of quantum mechanics. It is suited to problems that have discrete answers: combinations, itineraries, schedules etc. https://en.m.wikipedia.org/wiki/Quantum_annealing