Chen Jin's decision to produce learning machines was not due to a sudden impulse. Learning machines had great prospects for the future. It was true that the technological challenges presented were a little too advanced for the current level of technology on Earth. But, it was not impossible to realize his goal, it was only highly challenging. The AI chip, Deep Learning, big data, cloud computing, etc. needed to develop this learning machine was already present on Earth. The technological conditions were all there.
Why did he want to specially produce a learning machine that was a hardware product instead of developing a software product? It was because this type of learning machine required a built-in high performing AI chip! This was needed to build up a more advanced neural algorithm. The CPU performance must be good too. It was impossible to achieve the best learning results with an ordinary hardware platform.