DeepMind: Reinforcement learning suffices to create general AI
Researchers from Britain's DeepMind say that for achieving general artificial intelligence reinforcement learning is sufficient. Maximising rewards will help develop the machine's intellectual capacities, they argue.