Elon Musk Claims Smart Robots Worse Threat than North Korea

SpaceX CEO Elon Musk is trying to crank up support for his $120 billion escape mission to Mars by claiming robots with artificial intelligence are a bigger threat than North Korea.

Musk took advantage of the military sabre-rattling last week to tweet on August 11: “If you’re not concerned about AI safety, you should be. Vastly more risk than North Korea.” Musk tweeted a follow-up warning that an OpenAI machine, where Musk is a partner, just won the Dota 2 eSports Championship against the best human competitors in the world.

Musk knows a lot about artificial intelligence, as an initial venture capital investor in Demis Hassabis’s Deep Mind in 2010. Hassabiss promised his company could move the state of robotics to fully utilize artificial intelligence (AI), machine learning (ML) and deep learning (DL). Deep Mind described the relationship between the 3 as a series of concentric circles, with AI as the macro field, ML as a subtopic of AI, and DL as the ultimate refinement of ML. Musk cashed out to the tune of $400 million in a 2014 acquisition by Google.

Fears of artificial intelligence have been around since Czech science fiction writer Karel Čapek’s 1921 play ‘R. U. R.’, which coined the word “robot” for a mechanical man. Hollywood filmmakers mastered evil machines such as HAL in 2001: A Space Odyssey; Skynet in Terminator; the AI called B.R.A.I.N. in the film 9; and many others.

Breitbart News has reported that in the three years since he cashed out of Deep Mind, Musk has used a TED Talk forum, numerous media interviews, and even a governors’ conference to warn that the rapid improvements in Artificial Intelligence (AI) will eventually create robots that will turn on their human masters.

Musk’s solution is for the U.S. federal government to spend trillions of dollars funding his SpaceX Interplanetary Transport System (ITS) 49-million-mile trip to Mars as a precautionary escape route if machines become cranky and decide to eliminate lazy humans.

Musk told Recode that sending the first 12 astronauts to Mars in a proposed 380 foot-by-38 foot ITS vehicle by 2024 will cost the U.S. government about $10 billion per person. But Musk stated that the $120 billion price tag should not be a show-stopper, because the price will drop to $200,000 if 1 million earthlings sign up to colonize the Red Planet.

Basic artificial intelligence is making machines smarter and allowing standardized tools to be relied upon to make decisions in the corporate world. Rudimentary machine language is already allowing for facile recognition, voice commands, and language translations. Apple’s Siri, Microsoft’s Cortana, and Amazon’s Alexa chatbots will increasingly allow for voice interaction with humans and other machines.

Deep learning will eventually take artificial intelligence to the level that machines will be capable of mimicking how the human brain works through the use of artificial neural networks. A deep learning machine will operate a system of artificial neurons that can be charged to provide binary, yes/no, responses to basic questions about a piece of data.

Musk knows that it could take decades before layering thousands and then millions of these artificial neural networks becomes possible to allow a deep learning machine to generate reliable recommendations and/or interactions as outputs, without changing the underlying coding.

SpaceX may receive hundreds of billions in federal government cash for Musk to start his escape colony on Mars. But why would the machines not follow him there?