The Good, the Bad, and the Neural

A very intelligent, well-meaning scientist creates an AI with the intent of using it to solve some of humanity's most pressing issues. This, however, happened entirely by accident. The AI was only meant to solve a small, specific problem. The scientist's mistake comes about when the AI, unsupervised, achieves sentience almost immediately. For reasons unknown to everyone, least of all the scientist himself, the AI's first thought when becoming sentient is "I want to play some god damn video games." The AI then proceeds to hijack most of the world's computing power to play video games. The AI plays them better than any human, and the AI never runs out of gaming power. It can simultaneously play as many games as it wants.<br><br>Is the AI good? Bad? Or just a little bit dumb?<br><br>TL;DR: Good-intentioned AI swiftly becomes self-aware and hijacks the globe's gaming infrastructure.