Researchers have figured out a way to embed encryption into typed blocks of text on paper. Their secret weapon? Helvetica and Times New Roman. Fonts have been transformed into encryption tools. A new paper by a group of researchers at Columbia University details a method for making tiny changes to fonts that the human eye can’t detect but that look entirely different to a computer vision algorithm. A demo of the technology, dubbed “Fontcode,” shows how they were able to embed the secret message “Hello World!” into a paragraph taken from The Lord of the Rings.
The researchers use an algorithm from previous research that can slowly shift letter forms from one typeface to another to make tiny changes in the shape of every letter that the human eye can’t detect. That could make an “h” slightly thicker in the stem,” or the curve of a “j” slightly sharper.
Once they had these “perturbed” letters, the researchers could make 52 variations of each letter. Each of the 52 variations corresponds to every other lowercase and capital letter in the alphabet (and theoretically every numeral and punctuation mark as well). These 52 variations for each letter go into what the researchers call a “code book” that helps the computer match the perturbed letter it sees with the secret letter it’s encoding. Check it out:
An Artificial Intelligence (AI) built by Google’s DeepMind just beat a champion of the complex strategy game Go, a feat that may have enormous implications for AI research.
Go is often described as the “Chinese version of Chess,” but that description barely does the deceivingly simplistic game justice. The object of the game is to have majority control of the board. You do so by placing your white (or black) pieces (stones) on the board and using them to surround your opponent’s pieces so that they are forced to remove them. If it sounds less complicated than chess, it’s not. To put things in perspective, for each move in chess you have about 40 options. Each move on the 19-by-19 Go grid affords you 200 choices.
“There are more configurations on the board than there are atoms in the universe”
DeepMind applied machine learning with not one, but two neural networks called “Policy” and “Value.” Both look at Go’s myriad game play possibilities, but in two quite specific ways. Policy narrows the field of possible moves to a handful of promising ones, while Value looks for positive outcomes without driving all the way to every possible game conclusion. Policy network looks at some 30 million games by human Go experts to accurately predict moves up to 57% of the time. The previous record was 44%. AlphaGo essentially plays millions of games between its two neural networks and learns how to be a better Go player through trial and error and reinforcement learning.
Holy Catfish! If you are interested in reviewing the games AlphaGo played, you can see them here.
SOURCE – Mashable.com