Articles Blog

Artificial Intelligence Expert Critiques Sci Fi Movies

Artificial Intelligence Expert Critiques Sci Fi Movies

my name is Kristen Lennox and I’m a principal data scientist at beyond limits my job is to help humans and computers interact with each other in industrial context today we’re going to be looking at some Hollywood depictions of artificial intelligence and I’m gonna be giving my view on how similar that is to what’s going on in the real world today 2001 a Space Odyssey two astronauts on a mission to Jupiter with a computer in their ship supposedly the most advanced artificially intelligent computer that’s ever been created good afternoon how how’s everything going good afternoon mr. aymer everything is going extremely well so how is is described as working like a human brain except a human brain that never makes mistakes computers really don’t work like that fundamentally computers don’t think like we think even when they’re displaying what appears to be emotion I’m afraid it’s trying to make you more comfortable but it’s not reflecting what’s going on under the hood what artificial intelligence does is more task-specific so we teach a computer to do a task like drive a car but we don’t have a computer that makes decisions outside of that task there is one thing that houses the 9000 series has a perfect operational record excluding situations where there’s a hardware failure this is actually correct computers will do exactly what they’ve been asked to do so any decisions that hal was making that didn’t make sense to the pilots it can only be attributable to human error during the journey the computer house appears to be making mistakes this leads of the astronauts to attempt to deactivate how I know that you and Frank were planning to disconnect me and I’m afraid that’s something I cannot allow to happen it’s actually very difficult to design artificial intelligence that can potentially be dangerous to humans mostly because it’s very difficult to get machines to recognize when humans are in danger the film is definitely not an accurate depiction of the state of AI in 2001 but I’d say this is a solid see I can’t believe I’m having this conversation in my computer do you not conversation with me her a world where they’re artificially intelligent operating systems so the main character in the movie gets an OS named Samantha and they form a relationship I should tell you that I’m not in a place to commit to anything right now did I say I wanted to commit to you I’m confused Samantha is a very human character so not just mimicking human emotion but actually experiencing it I don’t know we’re looking at those people I fantasized that I was walking next to you and that I had a body which makes for a very interesting movie but is not the way that computers behave today we don’t actually know how to make a computer that lets things I want to learn everything about everything I want to eat it all up I want to discover myself that’s capable of growing beyond what we asked of it eventually the computers reached a point where they were much more interested in each other and much more interested in things that were not relating to the humans who created them what’s called artificial general intelligence which would be a true sort of conscious AI we don’t have a reason to believe it’s impossible but we don’t know how to get there and can we even do it with a with a current digital architecture or would we have to switch to something that’s closer to what our brains are which is you know an analog biological architecture but that doesn’t mean that we can’t get there it just means that we really don’t know when we’d be able to it’s a very interesting film it’s not accurate and it’s a lot of interesting ideas about how things could potentially evolve in the future I’d say B – ex machina are you building an AI I’ve already built one a billionaire tech genius who builds a humanoid AI robot and recruits a human – to test it the real test is to show you that she’s a robot and then see if you still feel she has consciousness so we don’t have a clear-cut definition of what consciousness is the definition that you know resonates the most with me is to say that consciousness is the experience of being yourself we don’t say that a thermometer knows what the temperature is because there’s no internal experience for a thermometer there is no reason to believe that our our current computers have any kind of internal experience would we know it if we got it if we built a very persuasive AI that could interact as if it had an internal life it’s very unrealistic to say that the first very human-like ki is going to fit in a human-sized box current supercomputers which are not capable of producing consciousness fit in buildings and are run by their own power plants there is some some hand waving in the movie about mythical quasi biological processing architecture that they’re going to use to be able to achieve this but again it’s it’s not based on anything that’s real and achievable right now it’s a very enjoyable movie it’s it’s a movie that sort of explores what does it mean to be human much more than it explores what does it mean to be AI and I would say if you were counting on this movie to teach you about AI you would not receive above it D – instead of creating an artificial intelligence he duplicated an existing one transcendence there’s an AI researcher who is dying in order to preserve him they upload his brain into a computer that he’s built I can’t describe it it’s like my mind has been set free they hook him up to the internet and suddenly he has access to all this information he never had he starts building things that humans can’t even comprehend we’ve made a breakthrough with the nanotechnology we can rebuild any material faster than before this movie is full of lies we can’t scan a person’s brain if we could we’d have to take your brain apart so don’t don’t try it on your own even with the best technology you can’t overcome physical limitations it only has so much storage space it only has so much processing space there’s only so much power to run it you don’t automatically get this sort of runaway increase in capability I think that there is a point to be made about taking the unknowable artificial entity and hooking it up to the Internet and letting it do whatever it wants financial markets but it’s unrealistic even in its depiction of the threat there the way the technology works is is a solid F wargames it’s basically about an artificially intelligent computer developed during the Cold War to sort of simulate global thermonuclear war and try to figure out under what scenarios would you maximize survival of of Americans a youthful hacker accidentally accesses this computer and asset two to play a game global thermonuclear war in trying to win this game the computer decides that the optimal strategy is to actually start an immediate preemptive strike against the Soviet Union and launch a bunch of missiles so the first takeaway is don’t ever give a computer the ability to launch your nuclear missiles that’s a terrible idea this computer which is very very smart in a very very inhuman way is doing exactly what it was programmed to do and it finds a solution and it goes to execute on it and it doesn’t understand that it doesn’t have all the context that really the best way to win a global thermonuclear war is to not ever start one in the first place it’s showing a computer that’s an antagonist but not a malevolent one it just doesn’t under stand Oh a plus a plus I love working when people make movies about computers they’re actually making movies about people we like watching stories about people or entities that are like us more than we would like watching you know this is the life of your toaster I think as long as people view it as entertainment and say that an artificial intelligence monster is no more real than a vampire or a zombie then it’s it’s good clean fun it’s actually a really exciting time to be working in artificial intelligence not because we’re building robots that have feelings but because we’re seeing a lot of really exciting things in healthcare in energy the the world of research in general has been completely revolutionized there’s so much more good to AI than than bad and that’s sort of the vision of AI that that I see moving forward [Music]

17 thoughts on “Artificial Intelligence Expert Critiques Sci Fi Movies”

  1. To make a human like AI, you would need to pre programme indefinite possibilities and give the AI all the tools to figure out the "correct" decision. An AI cannot learn to become a person, it can really only learn to do what task it has been told. It comes down to decisions and a computer cannot, at this time, make a complex decision like a human. And even if one could come up with it, the AI would need to grow like a human and make decisions based on based experiences and those experiences may effect unrelated decisions.

  2. She should have defined intelligence, too. After all, intelligence and consciousness aren't the same thing—surely insects are conscious, meaning that they are aware of their environs and aware of their awareness of their environs; but they aren't very intelligent. Perhaps she might have said that we have a type of self-awareness of the way others' self-awareness is a part of our own: I am aware of the fact that you are reading my words and am also aware of my awareness, just as I am aware of your awareness of my awareness. We are also beings who have capacities of memory, abstraction, imagination, speech, emotions, habits, and the like, that are far more complex than simply carrying out a task.

  3. The in-joke in the 2001 movie was the name of the Computer, HAL. By moving forward one letter for each of H A L you get I B M, the world's leading computer manufacturer at the time.

  4. I would love to hear her thoughts on the movie “I am Mother” about the robot who raised the small girl but ended up being evil.

Leave a Reply

Your email address will not be published. Required fields are marked *