Ex Machina Scientific Advisor: Real-Life Robots Will Sooner Have Ava's Intelligence than Her Body
The upcoming film Ex Machina explores a twisted version of the Turing Test, featuring an eerily human-like robot named Ava. But is Ava just a figment of Alex Garland's imagination, or are artificial intelligence experts close to realizing the singularity? One of the film's scientific advisors, Adam Rutherford, wrote a guest post in The Telegraph that essentially asserted that true artificial intelligence is not as far off as we might think, but an artificially produced human-like body is decades away at the very least.
First, he elucidates the film's explanation for Ava's human-level intelligence, and why he deemed it to be perfectly "plausible" during his consultation on the film: "Ava's consciousness is drawn from the BlueBook, the film's version of Google: her knowledge and behavior are generated by harvesting the teraflops of information we reveal about ourselves when we're using the internet. It's a brilliant explanation for the generation of her human-level intelligence, and it does have a root in current tech. We live in a world, after all, where every word you tap into Google, every purchase you make on Amazon and every Instagram snapshot you post reveals something about you, and about people in general."
He qualifies that artificial intelligence isn't likely to achieve any sort of "consciousness" anytime soon, but not necessarily because it's implausible, simple because AI researchers aren't treating consciousness as their primary goal. "Mostly scientists are building intelligent systems for specific functions. When Deep Blue beat Garry Kasparov at chess in 1997, it was a landmark in machine intelligence. But Deep Blue couldn't play a single round of noughts and crosses."
He then discussed the fact that, perhaps counterintuitively, Ava's physical prowess is much more of a stretch of the imagination than her intellectual prowess or self-awareness. "Ava's body... is probably decades from realization. Currently, scientists struggle to get robots to do things we find trivially easy: they can drive a car, but not actually get into one. Four billion years of evolution is a hefty head start." Although he concedes that science is making advancements in this area, with Boston Dynamics' Cheetah robot serving as a recent notable example:
Based on the progress of artificial intelligence thus far, Rutherford concludes that Stephen Hawking's warning that AI could spell the end for the human race was possibly "too dire." That being said, the film's implicit assertion that huge corporations like Google and Facebook would be the ideal candidates for creating dangerous AI might not be far off the mark, especially since Google purchased artificial intelligence company DeepMind last year.
"Ex Machina's Nathan Bates is a mix of Victor Frankenstein, Colonel Kurtz, Steve Jobs and Mark Zuckerberg; he is a man who aspires to put himself at the centre of creation. Ava's AI is drawn from his earlier creation, BlueBook, a simulacrum of Google. So guess who bought DeepMind for £400 million last year? And guess who also picked up Boston Dynamics and a suite of other robotics start-ups in the space of nine months? Google, one presumes, is acquiring these robotics and AI firms to help build evermore sophisticated technology in order to… well, we don't know."