Humans and Technology (05:26)

FREE PREVIEW

Professor Nigel Shadbolt recently co-authored "The Digital Ape: How to Live in Peace with Smart Machines." He discusses "The Naked Ape" by Desmond Morris. Our ancestors made stone tools millions of years ago—evolving the brain and muscular-skeletal structure.

Moore's Law (04:17)

Humans have difficulty recognizing technological progress. Computer processors have doubled in power approximately every 15 months since 1971. Shadbolt shows an example of IPhone nanotechnology.

Future of AI (03:53)

Humans have become super-specialized and interdependent. Engineers are producing advanced systems, but are not creating generally intelligent machines with self-awareness. Despite Stephen Hawking’s moral concerns, Shadbolt does not believe AI inherently threatens humanity—but inappropriate use of it may.

Early AI (04:01)

In 1978, Shadbolt began building expert systems to replicate aspects of human intelligence—discovering the difficulty of programming common sense into a machine. In 1997, IBM's supercomputer used processing power to defeat Garry Kasparov at chess.

The Semantic Web (01:53)

Shadbolt and Tim Berners-Lee put AI into the Internet, developing mark-up language to infuse meaning into websites. Supercomputer Watson became a Jeopardy champion.

AI, Biology, and Neural Networks (04:14)

Based on a human synaptic model, Shadbolt built neural networks that could evolve their weight and structure for two different robots. See Google DeepMind's algorithm learning to play Atari Breakout, given only sensory input but no rules.

Deep Learning (03:05)

Shadbolt explains how algorithms can identify correlations and patterns. A system has beat Go champion Lee Sedol, raising concerns of an AI takeover. AI devices will become companions, affecting social relationships. See a running robot using neural network controllers.

Preparing for an AI Future (01:25)

Machines are becoming more capable and powerful, but Shadbolt believes humans will always invent new jobs. We should understand the opportunities and dangers of the technology, and limit companies possessing data monopolies.

Credits: Can We Control AI: Asking the Right Questions (00:23)

Credits: Can We Control AI: Asking the Right Questions