DeepMind teaches machines to program Bitcoin, an important step forward in scaling

DeepMind Neural Programmer Interpreter Teaching machine learning programming is an important challenge for artificial intelligence. A paper from the Google DeepMind team , Neural Programmer-Interpreters , uses LSTM models and other modules to complete this challenge. Through training, DeepMind has enabled NPI to learn 21 programs, including simple addition and sorting, as well as complex 3D model standard conversion programs. NPI can continue to learn new programs without forgetting the programs it has learned, and has good generalization ability. We believe that this may change the survival of traditional software development companies, especially software outsourcing companies.

The increase in social acceptance of Bitcoin expansion and the growth in transaction volume have led to the growth of blocks, and the block capacity is about to reach the 1M upper limit. The debate over Bitcoin block expansion has been going on for a long time. The hard fork method of directly increasing the block capacity requires upgrading all software running Bitcoin, which may cause the collapse of the entire Bitcoin network. The Bitcoin community proposed to use the soft fork method of segregated witness to increase the block capacity in disguise. The developer team has made progress and it is expected to be released soon. Bitcoin expansion will completely solve the current inherent defects of Bitcoin and increase its application scope.
Technology Industry News: Baofeng TV releases the world's first VR TV, NVIDIA officially releases GeForce GTX1080 GTX1070 graphics cards, Siri 's creator will launch a new AI robot VIV next week , IBM opens up a new frontier in quantum computing, the public can try out IBM's quantum processor, Google's new patent replaces the lens with an electronic eye, which can correct myopia and take pictures, FliR Systems of the United States launches a new radioactive thermal imaging camera, a German company invents a balloon-shaped drone for indoor delivery, and in order to train AI to speak human language, Google lets it read three thousand romance novels.