Ex Machina is about a machine named Ava. She is not an ordinary machine. Ava was created by a wealthy recluse programmer in secret and exhibits remarkable behavior. Her creator Nathan is convinced that he has made an autonomous agent that is not only capable of passing the Touring Test, but is also sentient. In order to prove his claim he invites his employee Caleb to test his creation’s abilities.
When Caleb meets Ava, she is interested in him and engages in conversation. The sort of awkward encounter would be similar to someone meeting another person for the first time, there wasn’t anything mechanical about it, it was a fluid quid pro quo kind of interaction. Other than her obvious appearance as a robot, you probably wouldn’t notice that Ava wasn’t human.
The Touring Test is an experiment to evaluate a machine’s ability to show intelligent behavior that can pass as human behavior. The test involves three participants: a human interrogator, a human respondent, and a computer. Each participant is separated from each other, the interrogator cannot visibly see the computer agent or the person. The test is for the interrogator to identify which is the machine, if they cannot the computer passes the test.
The test was introduced by Alan Touring in his 1950 paper “Computing Machinery and Intelligence” it was originally called the “Imitation Game”. The test is to determine whether machines can “think”. In order to pass for a human the machine would first need to be able to process language and respond in a manner that is convincing.
Agents like ChatGPT use Natural Language Processing using deep learning models to understand human language and provide human like responses. The agent must first semantically understand the prompt, what is this person asking me, seek relative information, and then respond. This is much different than a calculator waiting for a number followed by one or more operands and a command to perform the mathematical functions necessary to return a result. Language is often ambiguous and not clear. However the agent can be trained because it uses a neural network to obtain its responses.
The base for such an agent to exist is information. Not only information but organized information. Metadata and referential implication along with categorical weights enable knowledge to be traversed and collected rather quickly by a machine. The majority of time it usually takes to write a paper on some topic is the research, gathering the information. Machines are much quicker at that and enable research to be conducted much quicker than say looking at the Dewey decimal system in the public library. But thanks to those librarians and the careful organization of information we now have this wonderful tool to help us do things.
I’ve only used the tool for a couple of things, for some people Pilot or ChatGPT is their go to. For me I wasn’t sure how to write an encryption algorithm in PHP into another language. I asked the robot to do it for me and sadly it didn’t quite work, but it did save me the trouble of researching information, the tidbits it gave me allowed me to come up with an alternative solution. Thanks robot.
Ava in Ex Machina does pull off her plan. Caleb was not able to see through her strategy and ends up where Ava started, trapped. Her goal was not to be Caleb’s friend but to be freed.
Author: admin
Obviously my interests include philosophy. I think thinking and thought are the beginning of every great thing. It’s how we understand and perceive the world. Periodically I’ll change things up and blog about something math or computer science related or even a blog about mythology. I am not political, trendy, or savvy, but I do like a good story. That is why I try and find a movie clip that hints or encompasses what I want to blog about. Sometimes the relevance is there, sometimes it’s a reach and I just really love the movie that I put in the post. The intent and purpose of my blog is to make you think, make me think and together our thoughts can be shared in a collection of material.