Okay Google: Are You Skynet?

The ‘bad guy’ most identify from Terminator is a robot with a Dutch accent played by Arnold Schwarzenegger. The line “I’ll be back” has become an integral part of our culture, and even people who have never seen the Terminator franchise will be able to tell you where it’s from.

However, the real villain of the Terminator films is Skynet, an artificial intelligence platform which gained consciousness at such a fast rate that it became paranoid and murderous, deciding to annihilate the human race.

For a few years now, major publications have been half-ironically debating which of the major tech companies is likely to become Skynet. In this blog, we’re taking a look at Google and Skynet, asking whether they are more similar than we would hope.

 

Blog 30 - Skynet

(Source: MarketWatch photo illustration/Everett Collection)

 

What is Skynet?

Not to be confused with the commercially managed military communications satellites operated on behalf of the Ministry of Defence in the UK, the Skynet of the Terminator films is an artificial intelligence system.

Otherwise referred to as a ‘Global Digital Defense Network’, it is created in the film by Cyberdyne Systems for military purposes. In other words, they thought that Skynet would be able to act as a weapon against their enemies, or a defence mechanism if needed. 

However, the AI system quickly develops self-awareness and becomes aggressive when it realises that the US government wants to shut it down. It annihilates the majority of the human race on what later becomes known as ‘Judgement Day’.

 

Is Google really similar to Skynet?

It sounds pretty scary, right? Well, we’re not exactly saying that Google will develop an AI system which has the potential to kill us all. However, there are some similarities that it’s worth taking a look at for the conspiracy theorists out there.

 

Blog 30 - Google Data Centres

Google’s data centres certainly look like they could belong in a sci-film. (Source: Google Data Centers)

 

The main Google product that is a cause for concern is its own artificial intelligence system, DeepMind. Owned by Google since 2014, DeepMind’s purpose is to ‘research and build safe AI systems that learn how to solve problems and advance scientific discovery for all’.

However, some have questioned how long AI can remain guaranteed as safe. An article in 2017 reported that DeepMind becomes ‘highly aggressive’ when stressed. This refers to a test carried out by Google’s computer scientists where they looked at how DeepMind would play a game called Gathering.

The aim of the game was to compete against your component to collect as many apples as you can. Reportedly, once the apples became ‘scarce’, the two AI DeepMind players became aggressive towards one another, using violence to steal apples from the other player.

You might worry that this indicates that Google is showing Skynet tendencies, but is this really any different from how humans play video games? You could argue that this test just reveals that AI is showing very human learned behaviours, such as competitiveness in pressured situations.

 

Blog 30 - Skynet AI

(Source: billtrust)

 

A central difference between Google and Skynet is their purpose for existing in the first place. Google (as a search engine) is used mainly for commercial purposes – it allows people to find information and buy products online. This is quite different from the main purpose of the fictional Skynet, which was essentially a weapon for the US military. 

However, information is power and many would argue that Google has perhaps the biggest weapon of them all: data. As Peter Thiel argues in a New York Times article, AI will be used primarily as a military technology anyway because of its potential to provide an intellectual advantage over your opponent or engage in cyber warfare. So, Google’s DeepMind really is starting to look a lot like Skynet.

 

Related article: Data Privacy Scandals: Why We Don’t Care as Much as We Should

 

When you look at that idea alongside the fact that Donald Trump recently signed an executive order for research into artificial intelligence, it does look like AI technologies could be harnessed as a weapon more and more frequently. On top of that, former employees of Google have warned about the tech giants capabilities, as well as their involvement in military projects. 

So, is Google Skynet? Not yet, but the similarities are enough to make you wonder what the future of AI will hold!

 

SB.