Do Machines have Consciousness?
Do machines have consciousness? One of last week’s speakers, Craig Martell, Head of Science and Engineering at LinkedIn, doesn’t think so. According to Mr. Martell, artificial intelligence (AI) is just a buzzword, purely named for its ability to raise money from the government or investors. The goal of AI is to build systems that act under uncertainty so humans don’t have to. He doesn’t believe AI machines will imitate humans, but are simply statistical models. He said, “It [AI] is tedious, boring, sweaty, and dirty. It’s not sexy.” I thought his view on AI was very practical and straightforward, stripping away the mystery behind AI. But I also think he was too dismissive about the lively and academic debate surrounding machines and their consciousness or lack thereof.
So first off, what is consciousness? Researchers and academics in neuroscience, psychology, philosophy, and computer science have tackled this question, but there are still many unknowns. However, as humans, we have a general understanding of consciousness. Conscious is what you experience. A Scientific American article beautifully described consciousness as “the tune stuck in your head, the sweetness of chocolate mousse, the throbbing pain of a toothache, the fierce love for your child and the bitter knowledge that eventually all feelings will end.” [1] Scientists describe human consciousness using three different levels: C0, C1, and C2. C0 is the subconscious processing of information, such as the processing used in facial recognition. [2] Most AI operate at C0. [2] C1 is the deliberate processing in response to certain stimuli. [2]. Finally, C2 is the ability to self-correct and explore the unknown. [2] Scientists argue that some AI can evaluate their actions and react accordingly, exhibiting some component of C2.
Currently, machines do not exhibit consciousness because they are missing one crucial feature – intention. Edith Elkind, a Computer Science Professor at the University of Oxford, said, “Machines will become conscious when they start to set their own goals and act according to these goals rather than do what they were programmed to do. This is different from autonomy: Even a fully autonomous car would still drive from A to B as told.” [2] We live in the era of Weak AI – computers can only simulate the brain, and a simulation of consciousness is not the real thing. [3] In 1980, John Searle, an American philosopher, conducted a thought experiment to prove this exact point. “The Chinese Room Argument” showed that “syntax is not efficient for semantics” – although the machine is able to translate Chinese into English, it does not mean that the machine understands either Chinese or English. [4]
However, some visionaries in the technology space, such as Stephen Hawking, Elon Musk, and Bill Gates, fear that as the machines become more intelligent they could take on lives of their own. [5] The Turing Test, developed by Alan Turing in 1950, says that if we can’t differentiate between a computer and a human, then a computer is intelligent. [5] So if a computer is intelligent, does consciousness follow? This is called Strong AI – the belief that machines possess the full range of human cognitive abilities, such as self-awareness, and sentience. [3].
Although AI researchers claim to be making progress towards Strong AI, we are still decades away. So for now, we can let science fiction, television, and film explore the idea. Just think of recent movies, such as “Her” and “Ex-Machina.”
If you are curious how the media is portraying Strong AI, here is the movie trailer for “Ex-Machina”:
Enjoy!
[1] https://www.scientificamerican.com/article/what-is-consciousness/
[2] https://www.livescience.com/62656-when-will-ai-be-conscious.html
[4] https://www.iep.utm.edu/chineser/
[5] https://medium.com/hummingbird-ventures/can-artificial-intelligence-be-conscious-e316c2ac4769
Users who have LIKED this post:
7 comments on “Do Machines have Consciousness?”
Comments are closed.
You share an interesting perspective that showed how the way in which people frame AI can influence the way we interpret the technology. Is it just purely algorithmic like the way Mr. Martell spoke or can we cast some psychologically-based frameworks to see AI with a more human-conscious like characteristic? It’s ultimately up to each individual to decide how they want to see and interpret AI, but I imagine that as AI gets further developed and adopted, we will likely see a trend develop that leans either towards or away from human-like AI.
Users who have LIKED this comment:
We might call artificial intelligence (AI) a buzzword – it is still great to see what advancements in science and technology have been made in a field that can be described with AI. It might not be a shortcoming of AI when we unmask its faults or incompleteness – it might be a shortcoming of our understanding when assuming AI systems would display human intelligence. We surely can appreciate the advancements in the combination of programming and statistics. We have to be careful when handing decisions to artificial systems that would require the inclusion of all the facets of the human mind.
Users who have LIKED this comment:
Before Craig Martell’s lecture I would probably have answered that, even tough right now machines do not have consciousness, AI’s development would have been able to provide them with it. I must admit that it somehow scared me, as it could someday overtake on human lives, as you pointed out. Nevertheless, if AI is all about statistical regression, as Martell simplified it to, seems like those tech visionaries’ fears, as well as mines, were totally unjustified. It also means that, despite machine decisions will become more precise with data training, it makes me doubt how well they will be able to react under unexpected events, and therefore Martell’s speech made me become skeptic on whether or not machines will ever have real consciousness.
Users who have LIKED this comment:
elauren, thanks for the post. Craig Martel’s “straightforward and practical” description of AI was one of the most insightful parts of this last week’s lecture as well. I was thinking of researching this idea, but you’ve done a great job of it already:)
From your research, what conclusions did you draw? Do you believe, as Craig does, that Elon Musk and others (Hawkings/Gates) are ill-informed and that AI will not develop consciousness? Or do you believe that it will ever progress to C2?
Users who have LIKED this comment:
Thank you! At the moment, I don’t think our technology is at the level to develop consciousness. However, I wouldn’t rule out the possibility in the distance future. I think it’s good that we discuss the matter, and think about the implications for the human race, but no one really knows the answer yet.
Great post @elauren!
Mr Martell’s AI break-down also unveils AI’s mystery for me as well. I think as you posted, well-defined consciousness level could also facilitate our understanding of AI and AGI in the future. Given human beings have been exploring the un-known and improving the intelligence machines ceaseless, I do believe in future breakthrough in strong AI! It’s just a matter of time 🙂
Great post elauren! I really liked the way Craig Martel approached the word AI as a buzzword used to glorify its real meaning: there is no real intelligence. Behind all the impressive functionalities of machines there is only a lot of coding, statistics and data. What makes people believe that “Artificial Intelligence” is, indeed, a form of self-awareness is the lack of knowledge on the field, on how machines work and what is behind each application or program. Having said that, I think the future exposed by Elon Musk where machines will actually develop consciousness is a bit far away from our reality.