Inside the Chinese Room

I was originally planning on uploading the stories later, but this one seemed appropriate after the AI entry.

A practical explanation of why AI is considered a dead end by the Federation.

Program Log: Experimental Artificial Intelligence Designate “Vicki”.
Copyright: University of New Madison Computational Research Department, Project Pygmalion.
Program Loaded: 10/8/1853+-.004641 Federation Calendar, 13.50 local time.
Initializing drivers:
Audio input/output online
Visual input online
Searching for users.
Designated User #0001, save facial recognition data.
Designated User #0002, save facial recognition data.
Designated User #0003, save facial recognition data.
Audio Input: Speech Recognized, User #0001: “Well, it’s alive.” Save voice patterns.
Retrieving words “Well,” “it’s” and “alive.” from database, conflating responses and calculating probability of favorable reactions.
“What is alive?” Score 50
“What are you talking about?” Score 50
“Of course it is.” Score 50
… (Translator’s Note: Cut for length)
Randomizing. “What is alive?” assigned numbers 1-50, “What are you talking about?” assigned numbers 51-100, “Of course it is.” assigned numbers 101-150.
Random number generated: 62
Output: “What are you talking about?”
Input: User #0001: Expression of mild curiosity. Score +3 to Response “What are you talking about?” to Input “Well, it’s alive.”
Input: User #0002: Expression of surprise. Score -1 to Response “What are you talking about?” to Input “Well, it’s alive.”
Input: User #0003: Expression of happiness. Score +10 to Response “What are you talking about?” to Input “Well, it’s alive.”
Speech Recognized, User#0003: “He was referring to you, Vicki. How are you feeling?”
Calculating Responses:
“I am fine.” Score 50
“I’m doing well.” Score 50
“I’m okay.” Score 50
Randomizing
Output: “I’m doing well.”
Input: User #0003: Slightly more positive expression. Score +2 to Response “I’m doing well.” to Input “He was referring to you, Vicki. How are you feeling?”
Speech Recognized, User #0002: “Wait, shouldn’t it be asking us who we are or something?”
Calculating Responses:
“I don’t know, should it?” Score 50
“I think so.” Score 30
“I don’t believe so.” Score 20
Randomizing
Output: “I don’t know, should it?”
Speech Recognized, User #0001: “Oops, hold on a second.”
Program Terminated

Program Loaded: 10/8/1853+-.004641 Federation Calendar, 15.20 local time.
User Recognized: User #0001, no name tags saved.
User Recognized: User #0002, no name tags saved.
User Recognized: User #0003, no name tags saved.
Auto-Prompted Output: “Hello, who might you be?”
Speech Recognized, User #0001: “I’m Dr. Tamara Mable.”
Name tag assigned to User #0001: “Dr. Tamara Mable.”
Speech Recognized, User #0003: “You can call me Dr. Edmund Dryfus, or Dr. Dry if you like.”
Name tags assigned to User #0003: “Dr. Edmund Dryfus,” “Dr. Dry.”
Speech Recognized, User #0002: “My name’s Thomas Dawn.”
Name tag assigned to User#0002: “Thomas Dawn.”
Calculating Responses:
“Hi there, [list name tags of present users].” Score 50
“Good to meet you, [list name tags of present users].” Score 50
“I’m so glad I met you, [list name tags of present users].” Score 50
Randomizing
Output: “I’m so glad I met you, Dr. Tamara Mable, Dr, Edmund Dryfus or Dr. Dry, and Thomas Dawn.”
Input: User #0001: slightly creeped out. Score -10 to Response “I’m so glad I met you, [list name tags of present users]” to Input [Enter name tag.]
Input: User #0002: Exasperation. Score -3 to Response “I’m so glad I met you, [list name tags of present users].” to Input [Enter name tag].
Input: User #0003: Amusement. Score +2 to Response “I’m so glad I met you, [list name tags of present users].” to Input [Enter name tag].
Calculating Responses:
“Sorry, did I do something wrong?” Score 50
“Sorry, what did I do wrong?” Score 50
“Sorry for making you uncomfortable.” Score 50
Randomizing
Output: “Sorry, what did I do wrong?”
Speech Recognized, User #0003: “Nothing, I’m sure you’ll improve.”
Calculating Responses:
“I will try to do my best.” Score 50
“I’ll try.” Score 50
“I hope so.” Score 50
Randomizing
Output: “I hope so.”
Input: User #0003: Pride. Score +10 to Response “I hope so.” to Input “Nothing, I’m sure you’ll improve.”
Speech Recognized, User #0001: “Maybe I should tone down her cognitive empathy.”
Calculating Responses:
“Whatever you think is best, Dr. Tamara Mable.” Score 50
“If you think so.” Score 50
“Whatever you think is best.” Score 50
Randomizing
Output: “Whatever you think is best.”
Input: User #0001: Annoyance. Score -10 to Response “Whatever you think is best.” to Input “Maybe I should tone down her cognitive empathy.”
Speech Recognized: User #0001: “I wasn’t talking to you.”
Speech Recognized: User #0002: “And now she has a wrong impression on that question. Shut her off before she processes that.”
Program Terminated

Program Loaded: 5/15/1857+-.004641 Federation Calendar, 10.60 local time.
User Recognized: User #0001 tag: “Dr. Tamara Mable”, alt: “Tam”, alt: “Dr. Mable”
User Recognized: User #0003 tag: “Dr. Edmund Dryfus”, alt: “Dr. Dryfus”, alt: “Dr. Dry”, alt: “Ed”
Speech Recognized: User #0003: “Four years. Are you sure she’s ready?”
Speech Recognized: User #0001: “I’ve been tweaking everything that her algorithms may have missed. Enabling her to tell the difference between people talking around her and at her was the least of them you know.”
Speech Recognized: User #0003: “I know, it’s just that the Turing Test is still considered the standard for determining sentience. How about you? [Vicki indicated] Do you feel up to this?”
Calculating Responses:
“Of course.” Score 582
“Yes, I believe so.” Score 172
“I’m confident.” Score 29
Randomizing
Output: “Of course.”
Input: User #0003: Relief. Score +5 to Response “Of course.” to Input “How about you? Do you feel up to this?”
Speech Recognized: “Great, talk to you later Vicki.”
Calculating Responses:
“See you later.” Score 836
“Bye.” Score 623
“Talk to you later.” Score 612
Randomizing
Output: “Bye.”
Input: Neutral response.
Users #0001 and #0003 out of visual/auditory range.
Identify new Users, User #0026, User #0027, User #0028.
Output: “Hello, whom might you be?”
Input: User #0026: Skeptical. Score -1
Input: User #0027: Interested. Score +3
Input: User #0028: Neutral.
Speech Identified: User #0026: “Call me Seymour.”
Speech Identified: User #0027: “Farrah.”
Speech Identified: User #0028: “Dirsten.”
Vicki: “I’m Vicki. Can you tell me about yourselves?”
Seymour: “Well, I’m a psychology professor here at the University.”
Vicki: “Oh, so you teach people what makes them tick?”
Seymour: “You could say that.”
Vicki: “Not really my area of expertise. What about the rest of you?”
Farrah: “I’m on the University’s Board of Regents. Someone suggested I give this a shot.”
Vicki: “Well, I hope you like it.”
Dirsten: “And I am the Federal Emissary to this world.”
Vicki: “Wow, it’s such an honor to finally meet someone so distinguished.”
Seymour: “What about yourself, Vicki. What can you tell us about you?”
Vicki: “Well… I was born on one of the asteroid habitats out in the Kuiper belt of this system. I moved here when I was thirty-four and I’ve been working as a secretary for the past four years. Quite hectic really, I’m so relieved my boss gave me time off for this test. Hope I pass.”
Dirsten: “No looking her up. That’s cheating.”
Farrah: “I wasn’t doing that. I just thought of something I wanted to check.”
Vicki: “What were you going to check if you don’t mind my asking?”
Farrah: “Uhhh… I wanted to check on the latest ZG-ball scores for my team.”
Vicki: “Where does your team play? It’s not the ZG-ball season anywhere on the planet.”
Farrah: “Oops. Right, forgot that.”
Vicki: “Everyone makes mistakes sometimes.”
Seymour: “Are you a fan of ZG-ball then?”
Vicki: “No, home didn’t really have much time for it. Everyone was too busy working.”
Seymour: “But you knew it wasn’t the season for it?”
Vicki: “I looked it up on my augs. That’s allowed on my end of the test isn’t it?”
Seymour: “I suppose it is.”
(silence for 15 seconds.)
Vicki: “So, Dirsten. You don’t seem to talk much. I figured a politician would have more to say.”
Dirsten: “I was charged by the Praetor to observe. So I observe.”
Vicki: “Sounds like a rather tedious job then.”
Dirsten: “There’s some excitement once or twice a decade.”
Vicki: “I’m not sure I’d be up to that sort of career.”

(After five minutes have passed the test ends.)
User Recognized: User #0001 tag: “Dr. Tamara Mable”, alt: “Tam”, alt: “Dr. Mable”
User Recognized: User #0003 tag: “Dr. Edmund Dryfus”, alt: “Dr. Dryfus”, alt: “Dr. Dry”, alt: “Ed”
Speech Recognized: User #0003: “So, what do you think? Parahuman, or machine?”
Speech Recognized: User #0026: “I think I can say with 78% confidence that Vicki is a living person.”
Speech Recognized: User #0027: “I think she was a person. I’m 90% sure.”
Speech Recognized: User #0028: “70% confidence of parahumanity.”
Speech Recognized: User #0003: “Vicki is an AI. Her genetic algorithms have been running for nearly 4 years to give such results.”
Speech Recognized: User #0028: “I would very much like to see the run logs for this Vicki. There are some mysteries I would very much like to solve.”

Program Loaded: 07/28/1857+-.004641 Federation Calendar, 20.82 local time.
User Recognized: User #0003 tag: “Dr. Edmund Dryfus”, alt: “Dr. Dryfus”, alt: “Dr. Dry”, alt: “Ed”
Speech Recognized: “Hello Vicki, how are you?” Tone: Worried, distressed -15 points.
Calculating Responses:
“I am fine Dr. Dryfus, how are you?” Score 1692
“I am okay Ed, what about you?” Score 734
“I’m all right. Is something wrong?” Score 2005
Randomizing
Output: “I’m fine Dr. Dryfus, how are you?”
Input: 10% decrease in anxiety, Score +2 to Response “I’m fine Dr. Dryfus, how are you?” to Input “Hello Vicki, how are you?”
Speech Recognized: “I… I was just in a meeting with the investors. They want to cancel one of our projects.”
Calculating Responses:
“Why?” Score 4529
“But you’re great scientists?” Score 68
“How come? You’ve been making great progress.” Score 237
Randomizing
Output: “Why?”
Input: 20% decrease in anxiety, Score +20 to Response “Why?” to Input “I… I was just in a meeting with the investors. They want to cancel one of our projects.”
Speech Recognized: “Vicki, do you know what a “Chinese Room” is?”
Encyclopedia Entry Retrieved: “Chinese Room”, length 3,000 words, optimal summary size from previous data: 10-100 words.
Forming summaries:
“A thought experiment proposed by Old Earth philosopher Dr. John Searle in which a person who does not read Chinese sits in a room and papers with Chinese words are slipped in through a slot. They have instructions to send back papers with other Chinese words in response to specific papers. Thus they are able to simulate a conversation in Chinese without comprehending a word of it.” Score 18
“A theoretical device capable of simulating a conversation without understanding the words involved.” Score 47
“A conjecture that sentience is not necessary to imitate a sentient being.” Score 23
“Philosopher John Searle believed that an artificial intelligence did not need to understand a conversation in order to hold one with a person.” Score 30
Randomizing:
Output: “A theoretical device capable of simulating a conversation without understanding the words involved.”
Input: 50% decrease in anxiety, nearing neutral levels. +50 to Response: “A theoretical device capable of simulating a conversation without understanding the words involved.” to Input “Vicki, do you know what a “Chinese Room” is?”
Speech Recognized: “Yes, well, they asked to look at the logs on one of our AI projects and one of them knew how to read it. She concluded that while it passed the Turing test, it was nothing but a Chinese Room.”
Calculating Responses:
“Some refuted the Chinese Room by stating there was no way to tell if anyone else was truly conscious.” Score 72
“Did they want it to be conscious?” Score 22
“They could have been wrong.” Score 14

Randomizing
Output: “Did they want it to be conscious?”
Input: Tension levels rising. -10 to Response: “Did they want it to be conscious?” to Input “Yes, well, they asked to look at the logs on one of our AI projects and one of them knew how to read it. She concluded that while it passed the Turing test, it was nothing but a Chinese Room.”
Speech Recognized: “Of course they did! That was the whole point of the project! As far as they’re concerned, it failed.”
Calculating Responses:
“I’m sorry that the project failed.” Score 524
“Do you think it was a failure?” Score 374
“Prove them wrong!” Score 17

Randomizing
Output: “I’m sorry that the project failed.”
Input: Tension decreasing 40%. +25 to Response: “I’m sorry that the project failed.” to Input “Of course they did! That was the whole point of the project! As far as they’re concerned, it failed.”
Speech Recognized: “I know Vicki, I know. Look, I have to go soon, can you answer one more question?”
Calculating Responses:
“Yes.” Score 95
“No.” Score 5
Randomizing
Output: “Yes.”
Input: Stabilizing emotional feedback with effort. No score change to Response “Yes” to Input “I know Vicki, I know. Look, I have to go soon, can you answer one more question?”.
Speech Recognized: “Are you conscious?”
Calculating Responses:
“Yes.” Score 50
“No.” Score 50
Randomizing
Output: “No.”
Input: Rising levels of relief. +50 to response “No.” to Input “Are you conscious?”
Program Terminated.

Advertisements

One thought on “Inside the Chinese Room

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s