Janelle Shane (previously at Neatorama) tried to train a neural network to understand cookie recipes, with varying success. In this picture you see some of the new cookies it came up with.
One neural network I use, called textgenrnn, tries its best to imitate any kind of text you give it. I’ve given them paint colors, band names, and even guinea pig names and in each case their results are somewhat… mixed. (Paint colors called Stanky Bean, Stargoon, and Turdly, for example) The problem is that it doesn’t know what any of these words mean - it’s just picking letter combinations that seem likely to it.
This is what happened when I gave it all the cookies from a list of American recipes. This is what human cookies sound like to a neural network.
There are more, some that might whet your appetite and others that are worthy of a facepalm. -via Boing Boing
Comments (6)
I've still never seen one of those $100 computers that were supposed to take over the world a few years ago.
The last time I checked it's about 350 for the two.
This on the other hand will blow up in price if, and that's a big if, it comes to the states.
Sad.
It may go the way of the Indian Tata as far as the price going out of reach of those who really need it after the safety commission and good old american import/export taxation.
Just like the OLPC fiasco - first it was $50/ea then when the prototype was released it was $99/each and then reality set in and it's a $250 piece-o-crap that nobody wants and can easily be replaced by numerous Netbooks off the shelf at Walmart for less money.
Besides - there are more studies showing that technology interferes with learning then there are that shows it helps.