Training AI: Reward is not enough

Scientist and author Herbert Roitblat responds to DeepMind’s “Reward is enough” paper with an alternate viewpoint. Read more


Google Just Launched New AI-Powered Algorithms

Google has gone live with its sentence compression algorithms, which use a neural network trained by dozens of linguists to directly answer users' search engine questions. While this process of supervised learning is complicated and expensive, tech compa...

Read more »