I was doing some reading/watching for an upcoming podcast on which I’ll be a guest, and I came upon a video of an analysis of a film. It’s a hard sci-fi/crime film which deals, in part, with artificial intelligence becoming “self-aware.” The video’s author thought the film’s implicit warning are applicable to real-life situations.
The concern is too hasty; there’s a catch: any program can be written to simply declare itself to be self-aware. Look, I just did it here. But why believe it in the first place? Machines can be programmed for deception (again, I did it here, too). Statistically, it’s much more likely that a machine is programmed to be deceptive in this regard than it is to be truly self-aware.
The problem isn’t limited to fictional narratives or machine-with-a-consciousness. I, as a real-life, rational agent, can declare myself a natural-born Plutonian citizen, vacationing on earth—reality bearing out this declaration is a different story.
All those scare-stories, like those of the Terminator universe, about computers becoming self-aware, are irrelevant since it’s a (probably) non-falsifiable claim, and its determination rests nearly entirely on our individual views on the philosophy of life and its origins. In many cases, that there could be a human(-like) consciousness inside a machine is also irrelevant, since human-like consciousnesses presume a minimum-level of empathy that prohibits the outright sociopathy and destruction. Those efficient, indifferent, yet self-aware machines are often portrayed without that trait. Thirdly, that the self-aware devices might hold a soul is not an issue if they’re trying to kill you. Ceteris paribus, with our instinctual self-preservation mechinisms, we kill anything that tries to kill us, human or not.
6 Comments
Yeah, those nameless fears are what sells, so our culture is inundated by such nonsense. There isn’t much entertainment in clear thinking.
Personification is a literary trope, but it’s taken to its ultimate conclusion in sci fi with machines. I guess imbuing machines with human like intelligence makes us wonder if intelligence could lead to being human, which is kind of silly. It really does come down to belief on origins.
Another way of looking at this whole conscious robot meme, is that it creates plausible deniability away from the programmers who programmed the robot to do diabolical things in the first place.
For example, who was it that decide to strap some C4 to the robot in the first place?
There will always be a Tyrrell hiding in some tower, insulated by his obscene wealth.
2nd thought: Of course, in the case of (Bladerunner’s) replicants, it is hinted that there is a biological component, so it is possible that biological systems (such as the instinctual self preservation that you mention) could be overriding their AI programming.
I only saw Bladerunner twice. and they were both a long time ago. I’ll need to watch it again, but I don’t remember the organic part of the replicant thing.
Jay,
I think both ‘replicant’ and the derogatory term ‘skin job’ hint at the bio-engineered component.
Here is a wiki, FWIW: https://en.m.wikipedia.org/wiki/Replicant