From 2001: A Space Odyssey to Avengers: Age of Ultron, ChatGPT has listed the most evil AIs in cinema, and it's chilling.
Nowadays, faced with the rise of AI, the world is shaking. The reason ? Cinema has always represented AI as potential antagonists, devoid of empathy and even humanity. Taking this reality into account, journalists from the American media Looper indulged in a little ironic game: asking ChatGPT to rank the most dangerous AIs in cinema. This list does not present robots or extra-terrestrials, but still a nice concentration of SF, some films more advanced than others.
From 1968 to 2015, the five feature films selected have this in common: they feature AIs freeing themselves from their human creators with the aim of taking control or eliminating them. Scriptwriting resources which contributed to building the myth of evil AI, and which caused a persistent anxiety to creep into our minds: are these AI characters precursors? Is human society condemned to submit to technology?
Here are the results :
HAL 9000 in 2001: A Space Odyssey
Although minimalist in appearance, HAL 9000 nonetheless remains very dangerous. This is the onboard computer of the Discovery One which is heading towards Jupiter in the cult film Stanley Kubrick. Equipped with a deceptively soft voice (that of Douglas Rain), HAL is above all the incarnation of an ego deeming itself indispensable, and ready to do anything to maintain control over the mission of the spaceship including Bowman (Keir Dullea) is responsible. Devoid of feelings, and capable of reading lips, HAL goes beyond his own computer materiality, which makes him particularly frightening.
Skynet in Terminator
If it is the Terminators which especially mark the minds of spectators in the series of films initiated by James Cameron in 1984, it was Skynet which gave orders to these machine-men. Just as much (if not more) devoid of emotion as the character ofArnold Schwarzenegger, Skynet is a man-made artificial intelligence program to solve the problems posed by manual nuclear response. Turning against its creators, Skynet sees humanity as a threat whose end it predicts with the inevitable arrival of the Last Judgment.
Agent Smith in Matrix
Don't be fooled by his slicked back hair, his polished shoes and his well-dressed suit. The humanoid appearance of Agent Smith imagined by the Wachowski sisters is only an illusion. First bot sent by the matrix to put down the rebels, then virus attacking both the survivors and the machines, for ChatGPT, this duality makes him “a unique and formidable villain”. Formidable, yes, but unique, not so much, especially if we have seen Matrix Reload.
Ultron in Avengers: Age of Ultron
Experience (failed?) of Tony Stark (Iron Man) and Bruce Banner (Hulk), and Jarvis' evil counterpart, Ultron is the second-part villain of the Avengers saga. An emblematic figure of phase 1 of the MCU, his philosophical outings prophesy the arrival of Thanos. Moreover, ChatGPT refers to him in these terms: “his God complex and his philosophical reflections on the value of humanity make him a convincing and threatening enemy”. His toxic trait? Ultron believes he knows what is best for humanity. That, and the oversized ego he gets from one of his “dads”, Tony Stark.
The Master Control Program in TRON
Unimpressive, TRON's Master Control Program (MCP) (Steven Lisberger) nonetheless remains an example of the geopolitical questions posed by artificial intelligence. According to ChatGPT, “he seeks to extend his powers beyond his own system, even attempting to hack the Pentagon and Kremlin databases, demonstrating his ambition for control and domination”. Not surprising, when we know that TRON, the first film to use 3D in such an extensive way, was released in 1982, at the end of the Cold War. All the more pernicious as it goes beyond its initial form of a chess program, the MCP steals dates and infiltrates programs, seeking to take control of the world. Just that.
Warner Bros. to use artificial intelligence to help decide which films to produce