reinforcement learning

The AI development group DeepMind says that all the core components of artificial general intelligence are already here, we just need to let them work.

Proud Parents

Reinforcement learning may be able to teach itself how to reach true intelligence.

Google's DeepMind Says It Has All the Tech It Needs for General AI

Competing AI systems tasked with setting a product's price eventually stop competing with low prices and instead gouge their customers.

Gucci Gang

Capitalism is just another game that the AI intends to win.

Two Pricing AIs Went Rogue and Formed a Cartel to Gouge Humans

A new 100-level platformer game designed by Unity's algorithms will test artificial intelligence systems' problem-solving skills.

Game Genie 2.0

The staggeringly difficult challenge is paying out $100,000.

Hellishly Hard New Game Is Specifically Designed to Confound AI

A new AI agent taught itself, using reinforcement learning, how to get dressed in the morning. The results could make animation much easier on artists.

New Friend

Somehow, this self-dressing algorithm is all of us getting up in the morning.

Having a Bad Day? An Adorable Video Shows AI Learning to Get Dressed

The car learns to drive like we do — through trial and error.

The Digest: A British Startup Is Using AI to Teach Cars to Drive Themselves

This AI learns by looking at mistakes as potential successes.

New Algorithm Lets AI Learn From Mistakes, Become a Little More Human

This is part of Musk's goal of creating smart but safe AI.

Elon Musk's Research Venture Has Trained AI To Teach Itself

It trains AI in challenging environments. 

Researchers Develop a New Algorithm to Teach AI to Learn — and How to Adapt

It may look like drunken parkour, but it’s actually the future of AI robotics.

DeepMind Has Taught an AI to Do Something Quite Remarkable

ai deep learning reinforcement learning curiosity autonomy

We’re one step closer to machines that learn and think like people.

Researchers Have Created an AI That Is Naturally Curious

DeepMind is learning by doing, just like a person would.

AI Is Being Taught to Solve Problems the Same Way Living Organisms Do

Latest Stories