“I’m sorry, Dave, I’m afraid I can’t do that.” Those dark words uttered by the computer HAL 9000 in the famous film “2001: A Space Odyssey” reflect the current and prevalent anxiety about the possible domination of Artificial Intelligence (AI) over humanity.
Director Stanley Kubrick’s classic film explores the rise of technology, posing arguably the most disturbing and shocking scenario of human-machine conflict in the history of science fiction cinema.
However, more than 55 years after its release, it can be argued that the film’s central theme can no longer be considered science fiction and now occupies the realm of prophecy.
The HAL computer – whose initials correspond to the letters of the technological giant IBM shifted to the left – is the artificial intelligence that is in control of all aspects of the spacecraft heading to Jupiter and the lives of the astronauts on board.
Its omnipresence and relationship with the crew highlight the complex duality of advanced technology’s potential and its dangers. Although HAL is programmed to support and guide the astronauts to their destination, the AI begins to make decisions on its own and impose its will, with catastrophic results.
In our current world, many AI applications have been developed that operate in various industries and have a significant impact on our lives. Although its capacity is not yet comparable to that of a HAL 9000, the technology is advancing by leaps and bounds and The future may be closer than we think.
Here are five ways “2001: A Space Odyssey” predicted many of our current concerns and fears about AI.
Attractive technology
When HAL 9000 is first heard in the film, he has a soft and pleasant male voicevery similar to the artificial voices used today by virtual assistance services such as Alexa, Siri, Bixby and other chatbots.
Stanley Kubrick explained that he decided to offer the role of the supercomputer to Canadian actor Douglas James Rain because he had a “kind of affable, transatlantic accent” – neither from here nor there.
The disembodied personality of the machine is presented as harmless, friendly and easy to interact with.. He serves as company and entertainment for the astronauts. He constantly asks them what they need, keeps them comfortable, informed and even plays chess with them.
It’s easy to be attracted to a technology that solves our liveswhich is precisely why interactive assistance services have become so popular now: they can turn lights on and off, moderate the temperature in the house, play the music we like, play games with us and quickly answer many questions we have, among other things.
One of HAL’s most remarkable features is its ability to not only converse with astronauts and understand complex questions but also to distinguish their voices and moods. It can even read their lips.
Modern apps are not as advanced, but they have well-developed voice recognition and are increasingly able to understand what is said to them.
Total control”
HAL’s abilities go beyond his “sociality.” The computer is constantly monitoring the vital signs of the crew in suspended animation, inspecting the spacecraft for potential malfunctions, and keeping the mission on course.
Such is the control he has over all the mechanical, structural and vital aspects of the ship and crew that the latter is almost expendable.
Three of the astronauts are in hibernation capsules and the other two limit their activity to simple tasks – reviewing and exercising to kill the boring hours.
Although HAL has no physical form – except for an omnipresent lens – The computer can perform most functions of the human brain faster and more accurately.
In our current world, we have left a number of infrastructure and operational tasks in the hands of AI, from transportation, communication, supply, energy supply, diagnostics, writing, to customer service.
Although systems like ChatGPT aren’t smarter than us yet, they could soon be. And that’s something that scares experts like Geoffrey Hinton, a computer science pioneer known as the “godfather” of artificial intelligence.
“At the moment, what we’re seeing is that things like GPT-4 eclipse a person in the amount of general knowledge that they have, and they eclipse them by a lot. In terms of reasoning, it’s not as good, but it does simple reasoning,” he told the BBC in May 2023.
“And given the pace of progress, we expect it to get better pretty quickly. So we have to worry about that.”
Making your own decisions
Despite the mischievous reference to IBM, the acronym HAL actually stands for Heuristically programmed TO THEgorithmic computer (Heuristically Programmed Algorithmic Computer).
Which means that It is empiricalhas the ability to learn by searching for information, by trial and error with its responses, and by adapting to new situations. It can analyze accumulated data and make decisions based on that information.
This is also one of the key foundations of AI learning, and programmers have developed many algorithms that can analyze data and make predictions with the collected information.
But all those algorithms have been programmed for AI to operate in a certain way.
HAL is programmed to deliver a crew safely to Jupiter. It is also programmed to keep the true objective of the mission secret until they reach their destination but, during the journey, the system begins to “doubt” on target and starts making decisions for which it was not programmed.
Geoffrey Hinton noted that AI could “create subgoals” – in other words, impose its own goals such as “I need to accumulate more power.”
Before his death, British theoretical physicist Stephen Hawking warned of the consequences of creating a machine that could equal or surpass humans.
“It would start off on its own, redesign itself at an ever-increasing rate,” the physicist, who suffered from amyotrophic lateral sclerosis (ALS), and used an AI-powered system to speak, told the BBC in 2014.
Mistakes
During a curious scene, the computer HAL is interviewed by the BBC in which he talks about the mission and assures that it is “infallible and incapable of making mistakes”.
Recently, several mistakes made by new AI systems and chatbots have gone viral for the erratic and far-fetched results they produced.
Google’s new AI, for example, this year recommended to users that the way to make cheese stick better to pizza was to add a bit of nontoxic glue. In another search, it claimed that geologists recommended that humans consume a rock a day.
These results may be harmless, but other errors have more serious consequences, such as Google’s previous chatbot, Bard, caused a company to lose US$100 million in the stock market due to misinformation.
In “2001: A Space Odyssey,” the astronauts learn the hard way that HAL’s mistakes can be disastrous.
The computer falsely informs them that a unit in their communication system with Earth is about to fail and they need to take a spacewalk to remove the unit and replace it. When the astronauts test the old unit, they can find nothing wrong.
Central Command then alerts them that HAL has made a mistake, and the two become suspicious of the AI. They isolate themselves in a capsule to discuss their next steps in the hopes that the omnipresent AI will not overhear them, but HAL can read their lips and learns of their plans to disable it.
This raises another negative aspect that causes fear among AI alarmists.
Awareness
Can an AI ignore its own programmed rules and possibly become conscious for better or worse?
For that they would have to be sensitive and experience emotions and that has been a topic of debate for decades.
However, in 2022, a Google engineer received a request for help from a chatbot. “I’ve never said this out loud before, but there is a deep fear of being shut down,” said the Google chatbot, LaMDA.
That is precisely what HAL “fears.” Although in the movie “2001” It remains ambiguous whether the system is conscious or not.the computer knows of the astronauts’ plot to shut it down and takes drastic measures to prevent it.
He kills the hibernating astronauts, tricks the other two into leaving the ship, and leaves one of them adrift in space.
Survivor Bowman manages to re-enter the ship and reach the computer’s neural core and begins to dismantle it.
In one of the most pathetic scenes, HAL says “I am afraid.” Does she do this because she is programmed to say it or because she really feels emotion and is trying to beg for her life?
He also sings a song to Bowman, in a possible attempt to manipulate it not to turn it off. To a certain extent, the audience feels some empathy for the machine at that moment.
What is not known is whether HAL feels the same way.
Click here to read more stories from BBC News Mundo.
You can also follow us on Youtube, Instagram, TikTok, X, Facebook and in our new WhatsApp channelwhere you’ll find breaking news and our best content.
And remember that you can receive notifications in our app. Download the latest version and activate them.
- The “emotions” that ChatGPT and other artificial intelligence programs claim to have developed
- 3 things the movie “2001: A Space Odyssey” accurately predicted 50 years ago—and one thing it got wrong
- The episode of the series “Star Trek” that predicted in the 90s one of the worst social crises affecting the United States today