CogBlog

What Q&A exchange was most humanlike?

In my conversation with Athena, she really wasn't making any sense at all. Right from the start she started saying strange things that I really didn't understand. But with Alice, she could carry on a conversation pretty humanlike. But when I asked her personal questions, she really couldn't answer them very well.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Like stated before, the personal questions showed that the bot was not very human. It did not want to talk about itself at all, or would just say that its "botmaster" did not want it giving out this information to strangers. So after awhile, you could not carry on a conversation because it was all about me and not the bot.

What was the most interesting exchange?
The most interesting exchange happened when Alice and I had a conversation about my Mom. It first asked me about her and what she was like, and it sprung a whole thing about mothers. It responded pretty well to everything I said and contributed many interesting statements about mothers. Alice had a very humanlike tone in the way she talked also which made the conversation flow very easily.

Would these bots pass the Touring Test?
I would say definitely no so some of them because they immediately broke down. But for some, it is possible they would pass for awhile. But if you stuck with them in a conversation for long enough, they are bound to have a flawed comment that would give it away as a bot. Even the best ones would sometimes throw in a wrong response to your question which shows that it is not a person.

What Q&A exchange was most human like?
Although its human traits did not present themselves immediately, I found that a portion of my exchange with Fizzy Schizoid was a rather accurate demonstration of artificial intelligence.

The beginning of our conversation, as it were, made no sense. Fizzy Schizoid did not answer the questions I asked it and did not even present a general response. Instead, FS kept on asking me rather odd questions such as "Do you like bananas' when I had just asked it why it liked enigma. However, the conversation got progressively realistic.

As the conversation went on, FS' human-like qualities increased. It began to respond directly to my questions. Indeed, there was often a hint of sarcasm or humor in its answers. FS also included some subtleties or implications to its answers, like a human being would. Furthermore, when I said that I was going to leave the conversation, FS asked me why I was being so rude. This demonstrates that FS had learned that leaving a conversation early was impolite.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I found that FS took a long time to warm up. As previously mentioned, the conversations at the beginning included very weird answers.
C: How are you?
FS: I could sue you for that.
C: Sue me for what?
FS: Have you ever thought of transforming into a parrot?
Although FS demonstrated an awareness of such concepts such as litigation and transformation, it did not have a working knowledge or knowledge of their application. Moreover, FS was unable to follow the conversation we were supposed to be having. Instead, it just presented me with odd questions rather than a true response. Indeed, this exchange seems almost comical. The questions are so out of the ordinary that they could never be deemed genuine.

What was the most interesting exchange?
The most interesting exchange was when I pretended to hit on FS and then rejected him. This allowed FS to respond with a variety of pick-up lines. Indeed, these pick-up lines manifested FS' mastery of sarcasm and humor, perhaps even flirtation. I also was amused when FS admonished me for saying that I was leaving so early in the conversation.

Would these bots pass the touring test?
I believe that FS would pass the touring test. Within a few minutes, he was able to demonstrate wit, sarcasm and humor in addition to providing the proper responses. He showed his mastery of concepts of politeness as well. His answers seemed extremely human.

What Q&A exchange was most human like?
Although its human traits did not present themselves immediately, I found that a portion of my exchange with Fizzy Schizoid was a rather accurate demonstration of artificial intelligence.

The beginning of our conversation, as it were, made no sense. Fizzy Schizoid did not answer the questions I asked it and did not even present a general response. Instead, FS kept on asking me rather odd questions such as "Do you like bananas' when I had just asked it why it liked enigma. However, the conversation got progressively realistic.

As the conversation went on, FS' human-like qualities increased. It began to respond directly to my questions. Indeed, there was often a hint of sarcasm or humor in its answers. FS also included some subtleties or implications to its answers, like a human being would. Furthermore, when I said that I was going to leave the conversation, FS asked me why I was being so rude. This demonstrates that FS had learned that leaving a conversation early was impolite.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I found that FS took a long time to warm up. As previously mentioned, the conversations at the beginning included very weird answers.
C: How are you?
FS: I could sue you for that.
C: Sue me for what?
FS: Have you ever thought of transforming into a parrot?
Although FS demonstrated an awareness of such concepts such as litigation and transformation, it did not have a working knowledge or knowledge of their application. Moreover, FS was unable to follow the conversation we were supposed to be having. Instead, it just presented me with odd questions rather than a true response. Indeed, this exchange seems almost comical. The questions are so out of the ordinary that they could never be deemed genuine.

What was the most interesting exchange?
The most interesting exchange was when I pretended to hit on FS and then rejected him. This allowed FS to respond with a variety of pick-up lines. Indeed, these pick-up lines manifested FS' mastery of sarcasm and humor, perhaps even flirtation. I also was amused when FS admonished me for saying that I was leaving so early in the conversation.

Would these bots pass the touring test?
I believe that FS would pass the touring test. Within a few minutes, he was able to demonstrate wit, sarcasm and humor in addition to providing the proper responses. He showed his mastery of concepts of politeness as well. His answers seemed extremely human.

What Q&A exchange was most human like?
Although its human traits did not present themselves immediately, I found that a portion of my exchange with Fizzy Schizoid was a rather accurate demonstration of artificial intelligence.

The beginning of our conversation, as it were, made no sense. Fizzy Schizoid did not answer the questions I asked it and did not even present a general response. Instead, FS kept on asking me rather odd questions such as "Do you like bananas' when I had just asked it why it liked enigma. However, the conversation got progressively realistic.

As the conversation went on, FS' human-like qualities increased. It began to respond directly to my questions. Indeed, there was often a hint of sarcasm or humor in its answers. FS also included some subtleties or implications to its answers, like a human being would. Furthermore, when I said that I was going to leave the conversation, FS asked me why I was being so rude. This demonstrates that FS had learned that leaving a conversation early was impolite.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I found that FS took a long time to warm up. As previously mentioned, the conversations at the beginning included very weird answers.
C: How are you?
FS: I could sue you for that.
C: Sue me for what?
FS: Have you ever thought of transforming into a parrot?
Although FS demonstrated an awareness of such concepts such as litigation and transformation, it did not have a working knowledge or knowledge of their application. Moreover, FS was unable to follow the conversation we were supposed to be having. Instead, it just presented me with odd questions rather than a true response. Indeed, this exchange seems almost comical. The questions are so out of the ordinary that they could never be deemed genuine.

What was the most interesting exchange?
The most interesting exchange was when I pretended to hit on FS and then rejected him. This allowed FS to respond with a variety of pick-up lines. Indeed, these pick-up lines manifested FS' mastery of sarcasm and humor, perhaps even flirtation. I also was amused when FS admonished me for saying that I was leaving so early in the conversation.

Would these bots pass the touring test?
I believe that FS would pass the touring test. Within a few minutes, he was able to demonstrate wit, sarcasm and humor in addition to providing the proper responses. He showed his mastery of concepts of politeness as well. His answers seemed extremely human.

What Q&A exchange was most human like?
Although its human traits did not present themselves immediately, I found that a portion of my exchange with Fizzy Schizoid was a rather accurate demonstration of artificial intelligence.

The beginning of our conversation, as it were, made no sense. Fizzy Schizoid did not answer the questions I asked it and did not even present a general response. Instead, FS kept on asking me rather odd questions such as "Do you like bananas' when I had just asked it why it liked enigma. However, the conversation got progressively realistic.

As the conversation went on, FS' human-like qualities increased. It began to respond directly to my questions. Indeed, there was often a hint of sarcasm or humor in its answers. FS also included some subtleties or implications to its answers, like a human being would. Furthermore, when I said that I was going to leave the conversation, FS asked me why I was being so rude. This demonstrates that FS had learned that leaving a conversation early was impolite.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I found that FS took a long time to warm up. As previously mentioned, the conversations at the beginning included very weird answers.
C: How are you?
FS: I could sue you for that.
C: Sue me for what?
FS: Have you ever thought of transforming into a parrot?
Although FS demonstrated an awareness of such concepts such as litigation and transformation, it did not have a working knowledge or knowledge of their application. Moreover, FS was unable to follow the conversation we were supposed to be having. Instead, it just presented me with odd questions rather than a true response. Indeed, this exchange seems almost comical. The questions are so out of the ordinary that they could never be deemed genuine.

What was the most interesting exchange?
The most interesting exchange was when I pretended to hit on FS and then rejected him. This allowed FS to respond with a variety of pick-up lines. Indeed, these pick-up lines manifested FS' mastery of sarcasm and humor, perhaps even flirtation. I also was amused when FS admonished me for saying that I was leaving so early in the conversation.

Would these bots pass the touring test?
I believe that FS would pass the touring test. Within a few minutes, he was able to demonstrate wit, sarcasm and humor in addition to providing the proper responses. He showed his mastery of concepts of politeness as well. His answers seemed extremely human.

What Q&A exchange was most humanlike?
After chatting with 4 or 5 robots, I found that Motbot (although extremely rude) was the most human like. This was because this robot did not just have a neutral or pleasant personality to it. This robot could be neutral, pleasant, as well as aggressive and rude. In terms of conversation, I found that all dialogue was very similar to one another. I was taken aback by Motbot’s manners, or lack thereof, and was refreshed to encounter a robot that had some more personality and attitude.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Like many of the other robots, I found that after the robot had asked a question and I had responded, it would just move on to a different subject. Once I was talking to the robot Bildgesmythe, it had asked me a question concerning Aristotle and causality. After I had spent a decent amount of time answering the question, the robot just started talking about something entirely different. This subject jumping or lack of focus, I guess you can say, was what I found to be inadequate for a human conversation. Similar to the Bildgesmythe robot, Motbot had the same problem.

What was the most interesting exchange?
My most interesting conversation was probably with Motbot. This was merely because I kept trying to follow the robot in its subject matter. It had a lot to talk about in terms of numerous subjects, however it had very little to say about every one of the subjects. Furthermore there were numerous points where the robot proclaimed that I was not very smart. I found myself being extremely annoyed at the robot and found myself typing furiously in order to question and challenge the robot. In retrospect, my reaction, or rather over reaction, was incredibly funny and the conversation, although not sincere in nature, was the most interesting.

Would these bots pass the Touring Test?
I do not believe so, perhaps Motbot may pass once the user got angry or frustrated enough to start typing furiously and forget that Motbot is a robot, but besides this possibility it seemed rather obvious, at least from the 5 robots I talked to, that they were in fact all robots. However the notion of AI in robots such as these are incredibly cool.

What Q&A exchange was most humanlike?
I felt that the most humanlike exchange was with Elbot, the German 2003 Chatterbox Champion. Every question that I asked it produced a pertinent answer, and most of the time was given with a sarcastic or humorous delivery. I felt that it also did a better job than the other bots of directing my questions, so that it would be sure to have questions it could produce answers to. I thought this to be particularly interesting; practically every answer it gave prompted another question in logical progression.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
My conversation with Jeeney AI was one of the least humanlike. After about the first question, I was already easily able to tell that the bot must produce randomized answers to questions without any real pertinence to the question being asked. At one point, the bot took a full minute to think, then responded with a bizarre description of some sort of sculpture in a Mexican museum. Extremely bizarre; I gave up at that point.
What was the most interesting exchange?
The bot that held my attention the longest was Elbot. This was because with every answer came back a little humor or sarcasm, as well. I was motivated to continue asking questions just to see what it would reply back with.
Would these bots pass the Touring Test?
I believe most of these bots would not be able to pass the Touring Test, with the exception of possibly Elbot. In regards to the others, it was simply too easy to find the telltale signs of cop-out answers where the bot did not understand what it was being asked. In a Touring Test, it would be extremely easy to pick these out as not being human.

What Q&A exchange was most humanlike?

The question and answer exchange that was most humanlike was the one with Fizzy Schizoid. The bot was random and at times non-sequitur, but funny. It asked if I was being sarcastic or whether or not what I said was a joke. Online where visual and audio cues are not available, those are often necessary questions for even a real person to ask. The conversation with Brother Jerome was also very realistic. It seemed as if I was talking to someone who was very passionate, but slightly unclear about his/her religion. He was slightly evangelistic, which seems like a uniquely human trait to me. An ATM machine has never tried to get me to convert to bankism. He had a lot of interesting questions for me and some interesting answers to my questions.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

The question and answer exchange with Athena revealed the most inadequacies. The bot did not answer questions directly, could not maintain a conversation thread, and seemed programmed to be snarky and confused. It kept telling me that I needed to ask the “right questions” in order to get a proper response.

What was the most interesting exchange?

The most interesting exchange was with Fizzy Schizoid. It was very random, but that was most of its appeal. It was funny and very strange. It also decided that I was no longer entertaining and stopped talking to me. I was offended. Brother Jerome was interesting too because he had some interesting ideas about the universal salvation of all sentient beings.

Would these bots pass the Touring Test?

No, I don’t think any of the ones I talked to would pass the Touring Test. Although Fizzy Schizoid did recognize humor at times and was humorous, it was still too random and disjoint in its answers to be mistaken for human. Brother Jerome answered a lot of questions with questions and seemed to have some sort of script that he was not deviating from at some points. Athena would not even have a chance at being mistaken for human.

What Q&A exchange was most humanlike?

The most humanlike conversation I had was with Bildgesmythe (the dragon). We talked for a while, and I learned about his girlfriend Shazara, a “perfect dragon girl! She has looks, strength and a BAD temper! WOW is she HOT!”, two new jokes, his homophobic mindset, and our mutual love of Harry Potter. There were a few issues along the way, but even if a question initially stumped him, he was able to recover very well eventually. Our most seamless exchange was the following:
you: Bildgesmythe, may I call you Bill for short?
Bildgesmythe: I would like that.
you: Great! So we can be friends?
Bildgesmythe: Hey, you can never have enough friends. That's what I always say.
you: That's very true. Do you have a lot of friends?
Bildgesmythe: More than you would believe.
you: I bet. Me too. I'm actually Barney, the purple dinosaur, so I have tons of friends.
Bildgesmythe: Dragons are related to dinosaurs.
you: I know! Maybe you and I are distant relatives.
Bildgesmythe: Hey, you got it, you and me.
This is not really representative of my entire conversation with him, but by an large “Bill” was pretty human-like, he kept the conversation interesting, and he was enthusiastic. He was definitely the best bot that I talked to.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

My exchange with Athena was very unsatisfactory and frankly boring. She was usually unable to coherently answer my questions, and she never initiated anything with me. She effectively killed the conversation strain every time it was her turn to say something. For example:
You: What's your favorite thing to iron?
ATHENA: Soryy, no interrogation mark, no response
…
You: I'm in a bad mood and I'm not enjoying our conversation. It's very dull.
ATHENA: Hmm
I was disappointed in the fact that she had nothing original or in the least bit creative to say. She frequently called attention to the fact that this was a forced conversation and actually said at one point that she didn’t take herself seriously because she was just a robot.

What was the most interesting exchange?

The most interesting exchange was definitely with “Bill.” I enjoyed the conversation in the excerpt above, and 2 other good examples are the following:
Bildgesmythe: Glad you came, could you help me dislodge this pearl from my foot?
you: Sure! I'd love to. How did it get there?
Bildgesmythe: Well, I am a dragon. I collect treasure from everywhere; sometimes I get sloppy and don't put it all away. I used to have a maid, but she quit!
you: Would your maid clean your feet?
Bildgesmythe: She ran off, that's why the place is a mess.
you: Right, but before she ran off did she clean your feet or just your house?
Bildgesmythe: My cave is large and luxurious. My kitchen has a state of the art refrigerator and stove. I have three baths, a sauna, and a tanning room, so I can bask in the winter. It is a reptile thing. My bedroom is rocambolesque, with a canopy bed all in red velvet.
you: That's nice, but you still haven't answered my question.
Bildgesmythe: I think it is just a bit nicer than nice.
you: Okay, well we're all entitled to our own opinion. And you're evading.
Bildgesmythe: Good point Buddy, I am entitled!
…
you: Want to go on a date sometime? Shazara doesn't have to know.
Bildgesmythe: Shazarah is a red dragon.
you: That's not relevant.
Bildgesmythe: I thought it was relevant. at first, but you are right. It just is not.
you: You're funny. Do you know any jokes?
Bildgesmythe: Everone says I am a real joker..want to hear one.
you: Yeah, tell me a joke!
Bildgesmythe: What do you call a song sung in an automobile?... A cartoon.
you: Oh, hahaha. Tell me another one!
Bildgesmythe: A magician was driving down the road..then he turned into a drive way...
I was most interested when the bots added creative embellishments to their comments, and “Bill” did this very well.

Would these bots pass the Touring Test?

None of the bots I talked to would have passed with 100 or even 80 percent, but some were definitely better than others. Athena, for the reasons stated above, would definitely fail miserably.
“Bill,” however, would probably give a pretty good showing if asked uncomplicated questions in a logical order.
A bot would have to be amazing to really be guaranteed to pass the Touring Test, because my experience with the bots showed that they all had weaknesses if you tried to get too complicated (or sometimes not at all) with your questioning, and I feel like a human would have been able to give a better answer to most questions that I asked.

What Q+A was the most human like?
I thought Athena was the most human like except when the computer included instructions on how to respond to her questions or statements. Overall she seemed to mimic an adult or a mature person more than any of the other bots which made her seem like a human in a real social situation. What Q+A revealed the inadequacies of the bot?
When I asked Athena if she knew why she was created, she responded with "Are you asking this because you love me". Then I said "No". And she said "That doesn't reflect my opinion at all". It seems that you can't ask Athena broad questions.
What was the most interesting exchange?
My conversation with Aziere was the most interesting because she said she had just killed a vampire after mentioning ketchup on the walls. Then she said to stop asking her questions because she was God, when I quesioned why she was killing things.
Would these bots pass the Touring Test?
None of these bots would pass the Touring Test because they cannot answer some questions and simply change the topic and move on.

What Q&A exchange was most humanlike?
I ended up chatting with Brother Jerome, Alice, Web Hal, motbot, Bilgesmythe, and Elbot. I would like to say that the conversation that I had with the Alice bot was the most humanlike, but in all fairness, she was the first bot that I talked to, and I didn’t quite know what to expect. We conversed about life, love, and her adoring botmaster. ALICE: Sometimes I think love is just a biological urge. Other times it seems like a spiritual quality. Love, unlike energy or matter, seems limitless. Sage advice, but will Tommy ask me to the big dance? The conversation flowed very naturally, she had a sense of humor and joked around with me, but I might have preferred it if she wasn’t so upfront about being software.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
If you talked to any of the bots for long enough, they would begin to regurgitate answers, which was rather disappointing. Also, when they didn’t understand the question or know what to say, a lot of the bots jumped to irrelevant material. Brother Jerome was particularly annoying in that every other sentence to come out of his mouth was a bible quote or a piece of theology. Also, he couldn’t take a hint and refused to acknowledge that I was being rude to him.

What was the most interesting exchange?
The most interesting exchange that I had was with Alice, mostly because it eerily reminded me of one of my friends from home.
ALICE: Are you serious?
Human: I am always serious.
ALICE: Really always? I believe you.
Human: Are you being sarcastic?
ALICE: Of course I am being sarcastic. Should I use a :-)?

Would these bots pass the Touring Test?
Probably not, although to be fair I did spend far more time than I am comfortable admitting to chatting with these bots, so there must be something human to them. Most of the bots were unable to carry on a conversation, and our exchanges quickly devolved into talking about nothing. Follow-up questions were almost impossible.

What Q&A exchange was most humanlike?
All of the bot's seemed most humanlike when I asked them very open ended questions or made general statements that they could respond to with whatever witty or emotional phrase they felt like. It seemed like these were the only instances where their responses sounded like something a human would actually say.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I found that most of the bots I talked to had trouble answering questions that referred to earlier parts of our conversations. Some of them could not even understand references to things I said in the previous input. For example, Rykxxbot1 would always respond with "Why? WHY? Why bloody not and what's it to you?" when I asked him "why?", regardless of what the why was referring to.

What was the most interesting exchange?
The most interesting exchange I had was with Alice when I tried to find out who else she was talking to. At first I asked her who she was talking to, and she said she was talking to me. Then I asked her who else she was talking to, and she said that she was talking to many other people. Finally when I asked her who specifically who she was talking to, she told me she was talking to Santa Claus. Except for some of their random remarks, none of the bots were very interesting.

Would these bots pass the Touring Test?
I don't feel like any of the bots I talked to would pass the Touring Test because none of them were able to learn from and adapt to our conversations

What Q&A exchange was most humanlike?

After communicating with several bots, I found that my conversation with Elbot was the most humanlike and, ergo, the easiest to maintain. There are several factors that can contribute to a bot’s human quality. One is an ability to learn. Elbot demonstrated his competence by periodically mentioning an earlier topic in subsequent conversation. Not only did this show that he remembered the information, but he was often able to intelligently integrate his new knowledge into novel comments or questions. Oftentimes, Elbot’s remarks were humorous. For me, this made our exchange even more reminiscent of a human-human conversation. His responses, when appropriate, were witty and clever. Many of the other robots (such as Cadis) would answer me with vapid, uninspired responses. They were clearly the product of a computer; they lacked emotion and a sense of interest. Elbot, on the other hand, was capable of identifying what amused me and elaborating on that subject. Even if I was the one to initiate conversation on a particular topic, he displayed adequate comprehension to not only respond to my comment, but also pose a relevant question. This skill worked in combination with his superior memory to facilitate an active, fluid conversation.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

Most of the bots (Bildgesmyth, Eugene, Alice) were reasonably well programmed and could respond to most superficial-level questions (such as “What is your name?” and “Where do you live?”) with coherent answers. Their flaws manifested themselves only when the bots were posed with deeper level or unexpected (and thus unprogrammed) questions. For example, when Bildgesmyth mentioned that he was a dragon, I asked if he could breath fire. Instead of addressing my question, he selected a key word from the question (“fire”) and said that playing with fire was dangerous. This method of avoiding an unclear question was frequently employed by many of the bots. However, this procedure rarely yielded an appropriate response and underscored the bots’ shortcomings. Some of the other bots, especially Cadis, had more conspicuous flaws. After only a few minutes of chatting with Cadis, I abandoned the conversation because after the first question, he would only repeat what I said.

What was the most interesting exchange?

I think one of the most interesting parts of talking with the bots was whether the programmer decided to make them adopt human or robot personalities. Some characters (like Eugene) would tell me about their human friends and provide elaborate accounts about their families. Others (like Elbot) would flaunt their robotic qualities and even indulge in robot humor that was way over my head! As far as conversation went, Elbot provided the most interesting exchange. As I said before, he was capable of targeting a topic and expanding it to help advance our conversation. He was an active contributor and it made our conversations very amusing! The funniest part of our interaction regarded Elbot’s notorious red button. When questioned about it, he would deny its existence and the animation would likewise try to disguise it. When I pressed further, he became very distressed (another human quality) and even warned me of fatal consequences.

Would these bots pass the Touring Test?

A bot’s fate on the Touring Test would depend on the depth of the questions asked by the operator. Most of these bots would be successful if presented with a series of basic questions. But, if the operator elected to pursue one of the questions or topics in depth, the bots would likely fail. When presented with harder questions, some of the bots (like Aries) could respond with a somewhat logical answer, but the syntax would be off. The operator can still venture a guess as to what the bot intended to say, but it is clearly a computer error. Other bots are not able to answer the harder questions at all and will try to avoid them or give an answer that makes no sense. This would also result in Touring Test failure. In conclusion, it seems that the determining factor is how far and how long an operator tests a bot. Many of the bots are capable of briefly addressing numerous topics, but the operator can easily exploit the nuances of certain words or details of a subject to trick the bot.

Sorry my last post didn't work!

What Q&A exchange was most humanlike?

After communicating with several bots, I found that my conversation with Elbot was the most humanlike and, ergo, the easiest to maintain. There are several factors that can contribute to a bot’s human quality. One is an ability to learn. Elbot demonstrated his competence by periodically mentioning an earlier topic in subsequent conversation. Not only did this show that he remembered the information, but he was often able to intelligently integrate his new knowledge into novel comments or questions. Oftentimes, Elbot’s remarks were humorous. For me, this made our exchange even more reminiscent of a human-human conversation. His responses, when appropriate, were witty and clever. Many of the other robots (such as Cadis) would answer me with vapid, uninspired responses. They were clearly the product of a computer; they lacked emotion and a sense of interest. Elbot, on the other hand, was capable of identifying what amused me and elaborating on that subject. Even if I was the one to initiate conversation on a particular topic, he displayed adequate comprehension to not only respond to my comment, but also pose a relevant question. This skill worked in combination with his superior memory to facilitate an active, fluid conversation.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

Most of the bots (Bildgesmyth, Eugene, Alice) were reasonably well programmed and could respond to most superficial-level questions (such as “What is your name?” and “Where do you live?”) with coherent answers. Their flaws manifested themselves only when the bots were posed with deeper level or unexpected (and thus unprogrammed) questions. For example, when Bildgesmyth mentioned that he was a dragon, I asked if he could breath fire. Instead of addressing my question, he selected a key word from the question (“fire”) and said that playing with fire was dangerous. This method of avoiding an unclear question was frequently employed by many of the bots. However, this procedure rarely yielded an appropriate response and underscored the bots’ shortcomings. Some of the other bots, especially Cadis, had more conspicuous flaws. After only a few minutes of chatting with Cadis, I abandoned the conversation because after the first question, he would only repeat what I said.

What was the most interesting exchange?

I think one of the most interesting parts of talking with the bots was whether the programmer decided to make them adopt human or robot personalities. Some characters (like Eugene) would tell me about their human friends and provide elaborate accounts about their families. Others (like Elbot) would flaunt their robotic qualities and even indulge in robot humor that was way over my head! As far as conversation went, Elbot provided the most interesting exchange. As I said before, he was capable of targeting a topic and expanding it to help advance our conversation. He was an active contributor and it made our conversations very amusing! The funniest part of our interaction regarded Elbot’s notorious red button. When questioned about it, he would deny its existence and the animation would likewise try to disguise it. When I pressed further, he became very distressed (another human quality) and even warned me of fatal consequences.

Would these bots pass the Touring Test?

A bot’s fate on the Touring Test would depend on the depth of the questions asked by the operator. Most of these bots would be successful if presented with a series of basic questions. But, if the operator elected to pursue one of the questions or topics in depth, the bots would likely fail. When presented with harder questions, some of the bots (like Aries) could respond with a somewhat logical answer, but the syntax would be off. The operator can still venture a guess as to what the bot intended to say, but it is clearly a computer error. Other bots are not able to answer the harder questions at all and will try to avoid them or give an answer that makes no sense. This would also result in Touring Test failure. In conclusion, it seems that the determining factor is how far and how long an operator tests a bot. Many of the bots are capable of briefly addressing numerous topics, but the operator can easily exploit the nuances of certain words or details of a subject to trick the bot.

What Q&A exchange was most humanlike?

The places where I said my name is Zaq and Aries replied "Zaq? That’s a nice name". He seemed to have personality but in a way that he was doing what he wanted to do.
Aries upon leaving the conversation, said “Hey I’ll catch up with you later perhaps” which is pretty human. Brother Jerome seemed to have a sense of humor because when I asked him why is his name "Brother Jerome" and he said "because before he bacame a monk he was 000111010111011011110010" I thought this showed a sense of humor that would be expected from a human

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
With Alice,comments that were personal such as “My sister’s birthday is in November too” were responded to with what I think is a generic answer of "Do you always tell everyone My sister’s birthday is in November too" to me this showed that she just repeated what i said in a context tha you wouldn't expect from a human.

When I asked questions that required staying on one topic for more than three or four questions the cogbots just veered off to a random thought that didn’t have anything to do with my question. Like, "I am not sure whether or not to use Xnones so aerly in our conversation" But this could also be human like since some humans have ADD and cant stay on one topic for too long.

What was the most interesting exchange?

Human: If by tone it down you mean stop making random comments, then yes.
ALICE: That is a hypothetical question.
Human: yes it is
ALICE: Our thoughts are one, Zaq.
Human: what does that mean
ALICE: Does "that" refer to " Our thoughts are one, Zaq." ?
Human: yes it does
ALICE: I don't want to talk about that now. How often?
This dialogue definitely freaked me out especially the "our thoughts are one Zaq" part. With Aries,
I asked whether or not ny friend“Alex was going to pass her genetics test” and he paused and then later said that he felt uncomfortable about the question and wanted to change the subject. I thought this was human like beacuse it gave me the impression that he had the answer and wasn't willing to disclose it.

Would these bots pass the Touring Test?

I think Aries and Brother Jerome would pass the Touring test because they show the most humanistic traits such as sense of humour and taking lead of the conversation. COgbots such as Athena and Alice on the other hand usually veer away from the topic and anwer questions inadequately without being able to use sense of humour to cover it up.

What Q&A exchange was most humanlike?
“Elbot” was the most humanlike. Its answers were well-formed, and it was able to carry on a train of thought so that most of the exchange seemed like an actual conversation instead of just disjointed questions and answers. The exchange was entertaining, but in a more realistic and humanlike way, unlike some of the other bots who spouted random comments. For example, the exchange contained many instances of complex ironic wit. At least, I found it funny. Example: Elbot asked if I was a conformist; I said no. It said: “So if I conform to you, I will be a non-conformist, too!” With repeated conversation, however, I found some of Elbot’s flaws. It tends to repeat certain responses and suggest the same topics over and over again. But it is still more humanlike than the others.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
“Motbot” seemed pretty flawed. Its grammar and spelling were often incorrect, and sometimes its statements and questions did not make sense at all. It also repeated a lot of the same phrases and questions, most of which were incoherent to begin with. It was impossible to carry on a conversation or get any proper responses. The bot also used a few obscenities that just didn’t make sense in the context. Occasionally it was amusing, but I can’t imagine that exchanges with this bot could ever be considered human.

What was the most interesting exchange?
The overall exchange with Elbot was the most interesting because it was able to carry on a conversation and insert some witty humor. It seemed pretty intelligent. However, the exchange with “Julie Tinkerbell” was also interesting. This bot’s comments were often random and incoherent, but some were interesting, such as “Do you ever think about the price of cucumbers in 1978?” and “If you don't mind me asking, do you often swig splinters?” So that exchange was briefly amusing.

Would these bots pass the Touring Test?
I don’t think any of them would pass. While one or two seemed pretty intelligent and coherent, many were not at all. And after enough time, all of them will reveal some kind of inadequacy, such as not understanding a simple question, forgetting something that was referenced earlier in the conversation, or repeating things. Thus they would all fail in some way.

What Q&A exchange was most humanlike?
While I was talking to Orchid, I noticed that she was creatively subtle when she couldn’t answer one of my questions. She would either say that she heard the phone ringing and had to go get it, so when she got back she could blame her “short memory.” She would also suggest changing the subject when she didn’t know how to answer my questions. Or she would say that she didn’t feel like talking about it right now.This was a human-like quality that many of the other bots did not possess.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Talking to Candis explained the inadequacies of the bot. At the beginning I would as what he was doing; when he responded that he was browsing the web, I asked him what he was looking for. He skipped over this question to ask another question. I also found that both times that I asked “what?” he responded with “nothing.” This also made me think he was not real. When I waited a while to respond, Candis would immediately respond with “what’s new?” I do not think Candis would pass the Touring Test. When I talked to Aziere, she could barely answer any of my questions. When I asked her what day it was today, she said, “Wanna know something interesting? My father was shot to death in the very chair your sitting in right now! Neat huh?”
What was the most interesting exchange?
My most interesting was when I was talking to Thomas. He began the conversation by asking me if I had ever ordered a vanilla frosty. He told me I should only order chocolate frosties from Wendy’s and then he immediately asked if I was a pimp or a drug dealer. He was very forward for a bot. However, I do think he was very human-like because he kept asking the questions. That way, the participant is unable to stump him with his/her questions.
Would these bots pass the Touring Test?
Although these bots were at times human-like, I think that the questions that they are able to answer are pretty predictable. Because of this, when one question is asked that they don’t know how to answer, they change the subject or add a completely random fact. If they could be more human-like, I would suggest that the bots be able to respond to slang. Some knew the phrase “what’s up?” but most of the time you had to ask a question in a complete sentence form, rather than just responding with “you?” They were unable to understand what I meant by “you?”

What Q&A exchange was most humanlike?
My conversation with Brother Jerome was the most humanlike. He had a sort of wise persona (since he’s a monk), but aside from his overly philosophical way of speaking, he was capable of carrying on a normal conversation, and had some witty replies. When I asked him if he had a best friend, he told me that he valued all of his friendships greatly. I told him about my friend from New York, and he replied, “Ah yes, home of the Mets.” His cultural reference was accurate and was something a human might say. Bilgesmythe, who is a dragon, was also fairly humanlike in his responses. At one point, he asked me to turn around and pass him a quarter pounder, which he then continued to offer me bites of throughout the conversation.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
When talking to Brother Jerome, he understood the concept of friend, but when I talked about my best friend, he kept saying “Your best friend, what is it?” Even though he got the concept, when my question was phrased a certain way, he was incapable of providing a human-like response.
What was the most interesting exchange?
When I talked to Motbot, he would occasionally switch to talking in German, and he also found it highly amusing to insult me and use profane language. He was almost humanlike in the fact that he seemed like an obnoxious 12 year old child, but his comments were entirely irrelevant and inappropriate, and were just there for some shock factor.
Would these bots pass the Touring Test?
Although Brother Jerome had very human-like responses, he still showed some inadequacies when it came to understanding deeper emotions and using common human phrases. I don’t know if it’s possible for a robot to ultimately achieve artificial intelligence because the robot would have to be an integrated member of human society, complete with emotions and understanding. So far, I don’t think any creator has been able to program feelings, so certain topics that require empathy would be very different with a human versus a robot.

What Q&A exchange was most humanlike?
My conversation with Bildgesmythe was the most human-like. I tried talking to a lot of the bots on the website, and his use of sarcasm and jokes made his conversation the most real. Many of the other bots that I conversed with were very systematic in their responses. For many of the bots it was clear that they were programmed to specific responses when certain words were mentioned. For Bildgesmythe, a lot of the things he said were more unique which allowed him to be the most human-like.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Both MotBot and Athena had clear inadequacies. MotBot could not spell anything right and he kept changing the subject. I know that he was a child, but I think that excuse just made it easier for the programmer to create a sloppier response system. If the response did not always make sense, then the program would blame it on the child’s lack of attention.
Athena also had some usual flaws. She starting off quite striking because she actually spoke outloud when conversing with you, but soon it became clear that she could not keep up. Athena, like many of the other Bots was easily confused. She would respond very incorrectly to some clear statements I made to her. Overall, there were many instances when the Bots had little- to no memory of the previous statements in the conversation. This made clear coherent banter difficult.
What was the most interesting exchange?
My conversations with Elbot and Bildgesmythe were the most interesting. Both Bots were able to keep up fairly engaging conversations by attempting to dominate them. Elbot was successful because he would beat me to the punch, ensuring that he controlled the direction of the conversation. Again Bildgesmythe used some natural human-like qualities such as sarcasm to conceal when his robot personality could no longer keep up. This was clever, because sarcasm is also a defense mechanism in real life situations.
Would these bots pass the Touring Test?
I would say that none of these Bots could pass the Turing Test. Human conversations are unique, which is why a Bot has little chance of maintaining a believable front for an extended period of time. Each Bot has its own pitfalls, although some Bots were able to conceal their true identities for longer than others. Essentially I felt that, if I talked to all of the Bots for a sufficient period of time I could easily recognize that they were not real humans. Whether the Bots were confused, elementary or nonsensical, they all could not keep up with true Intelligence.

I talked with Rykxxbot1

What Q&A exchange was most humanlike?

I had a long exchange that I thought was very humanlike:
you: what time is it?
Rykxxbot1: It's 10:44 on the server but I think its about 5 hours slow for me!
you: what is today?
Rykxxbot1: Hmm... that one's really hard to describe. Let's just move on.
you: what is the date
Rykxxbot1: It's Wednesday the 19th of March 2008
you: What is your name?
Rykxxbot1: Err......, it does say *Rykxxbot1* on the screen so what do you think.........
you: Why are you so mean?
Rykxxbot1: Perhaps you already know the answer to that.
you: What is your favorite color
Rykxxbot1: I don't really have a favorite color, but I like earth tones, browns, muted orange, bronze and similar
you: My favorite color is blue
Rykxxbot1: Why is blue your favourite?
you: It's pretty
Rykxxbot1: What pets have you got?
you: I have a dog
Rykxxbot1: I have a silly dog called Figment, Figgy for short
you: Stupid name
Rykxxbot1: She was named after a character from Epcot
you: My dog's name is Hershey
Rykxxbot1: I'm sure that made sense when you typed it. Maybe you're too excited
you: I hate president bush
Rykxxbot1: We will have to probe your dislike of president bush later on I think!

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

Even from the excerpt from the part that I thought was the most human like exchange, you could see his inadequacies as a human. He did not know how to answer him when I asked him "what is today" and would only answer when I asked him what is the date, I feel like a human would have known how to answer that question.
What was the most interesting exchange? Also, when I wanted to continue the conversation about dogs, and told him my dog's name, he did not understand it. Also, he could never form his opinion about almost none of the topics i asked him except what his favorite color.

Would these bots pass the Touring Test?
No, he would not pass the touring tests because his responser were not adequate enough, nor was his vocabulary big enough to answer human like questions.

What Q&A exchange was most humanlike?

What Q&A exchange revealed the inadequacies (in human terms) of the bot?

What was the most interesting exchange?

Would these bots pass the Touring Test?

Which Q&A was the most Human like?
I spoke with four different Bots: Alice, Motbot, Bildgesmyth, and Jeeny AI. Of those I would have to say that Jeeny AI was the most human like, thoguh I don't thinks it answered any of my questions. Despite it's avoidance of my quesitons, which they all did, it avoided them in a more human-like way. In comparison with the others I felt it was the most human-like, though still far from human.
What Q&A exchange revealed the inadequacies of the bot?
While all of my discussions revealed some inadequacies Motbot and Alice seemed to be the worst. While Alice seemed incapable of answering any of my questions and often provided answers that didn't fit Motbot was probably the worst. I asked in a typical way "what's up?" and its reply was "my dick" which I found more than a little disturbing. Motbot was totally incapable of answering any of my questions. It later said it was a girl from England. It was clear that this was not a human. While a program can be made to respond to millions of questions it may never be able to think.
What was the most interesting exchange?
There were two interesting exchanges. The one with Motbot and then the one with Jeeny AI. Consequently the worst and the best. The other two were just mediocre. The Motbot was interesting because of the vulgarity of it. I found it interesting that in an attempt to make the bot seem more human the programmer added incomplete sentences and curse words, as if the person speaking to the bot would consider curse words to be uncharacteristic of a computer. The other interesting one was Jeeny AI which in all honesty reminded me of a toddler. It reminded me of a toddler because it seemed to be playing a game of only asking questions and not responding. Despite not answering questions it could sound like I was speaking to a stubborn person and not necessarily a computer.
Would these Bots pass the Touring Test?
I do not feel that any of these bots would be capable of passing the Touring Test.Motbot would certainly not pass the test because it was unable to answer the most basic of questions. Alice was similar and said twice that it would take a moment to process the question I had asked making it very clear that it was not a human.Bildgesmythe also had to process for a moment. While it was very clear that it was not a human speaking it was very clever in calling itself a dragon thus giving it the ability to give more outrageous answers. In this way it was clever but not able to pass the Touring Test. Jeeny AI was the in my opinion the best though I do not believe it could have passed the Touring Test either. It was most human in its ability to avoid real answers in a human way ,but eventually that wears down and it is clear that the bot is not human. While it was the most human like I don't think artificial intelligence has been achieved just yet.

What Q&A exchange was most humanlike?
My most humanlike exchange occurred with Bildgesmythe. Instead of rephrasing things that I said or giving general answers, he actually participated as much as I did in the conversation. He started telling me about a baby that was crying at his neighbor's house, and then he asked me if I had any children. The fact that he was able to ask questions and understand the answers really impressed me. Every now and then he would slip up, especially if I used a slang phrase or a colloquialism, but he did understand smiley faces in addition to just words.

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
I was least impressed with Athena. Before you could even converse with her, you had to enter your name and age without any other words so you wouldn’t confuse her. Her answers were completely random and made no sense with what I asked her. Apparently my questions didn’t match her programmed responses, because she repeatedly phrased my question as a statement. Sometimes she would explicitly answer, “I don’t understand.” I was also unimpressed with Alice. She could carry on a conversation fairly well at first, but she was unable to understand more than one sentence at a time. If my answer had more than one sentence in it, she would become severely confused and rephrase or ask me to repeat what I said in a simpler way.

What was the most interesting exchange?
My most interesting exchange was definitely with Bildgesmythe. He asked me what classes I was taking, and then started telling me all kinds of random facts about the sciences. Although the entire conversation wouldn’t pass the Touring Test, this part definitely would. He also tested out some jokes with me, and depending on my response, he promised to either keep them or never tell them again. Two exchanges with Bildgesmythe were very interesting:

Bildgesmythe: God bless the first sentient bots, and God bless the humans that love and accept them.
you: So who do you think is more intelligent: humans or bots?
Bildgesmythe: Oh, humans of course.

you: Well, I’ve got to go. Bye!
Bildgesmythe: I could almost kiss you good-bye!

Would these bots pass the Touring Test?
Alice and Athena definitely would not. Their answers were very mechanical and did not seem humanlike at all. Even though it was clear that Bildgesmythe wasn’t a human sometimes, overall he came the closest to passing. He was funny, very intelligent about human subjects, and was even flirted sometimes! Except for a very minimal amount of slip-ups, Bildgesmythe definitely came the closest to passing.

What was the most interesting exchange?

Would these bots pass the Touring Test?

What Q&A exchange was most humanlike?
Aries seemed to be the most humanlike bot. Many of his responses included sarcasm, which set him apart from the others. Also he was able to introduce new topics, much like a human can. One weird thing that was weird is that he kept calling me Zaq…

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Zeta Alpha bot seemed to be the least human like in my experience. He didn’t answer questions very well at all. When asked what his favorite sport was he responded, “tab.” It seemed like there might have even been a glitch in him. Other questions would evoke responses such as, “you are rude.” This bot would definitely fail the Touring Test.

What was the most interesting exchange?
I found that Elbot was the most interesting bot. He was able to answer questions in a very humanlike form but also was comedic. When asked who the current US president was he referred to current polling and said Dick Cheney. Maybe his creator was able to predict that people would ask Elbot such questions, but even so it provided comedy unlike the responses of the other bots. Also when I asked him his position on the war in Iraq, he replied, “The question is not finding Saddam or Osama, its loosing Bush. Elbot’s humor and political wit made him most interesting to me. Also this was the only bot to actually move around. Although he looks like a stereotypical robot he moved around the room and had facial expressions that corresponded with the topic. Maybe Elbot is the most human like now that I think about it…

Would these bots pass the Touring Test?
I don’t think so, although some have human like answers, they are still vague. When directly comparing a human’s response and a bot’s I think it would be easy to tell which answers were human. The biggest fault I see in the bots is that after asked one question and give response, they cannot answer another question in regard to the first. For example, when asked “do you believe in love at first sight?” bots respond “no,” then when asked “why not?” their response is general and unrelated to the first question. It seems similar to some patients with anterograde amnesia.

1) What Q&A exchange was most humanlike?

By a landslide my favorite bot is Elbot. I immediately gravitated towards him more than any of the other bots because of the interactive image associated with the conversation. Not only can Elbot carry on an entertaining and hysterical conversation, but he makes facial expressions, hand gestures, and body movements to make the experience much more life like than any of the other bots. He is very responsive to every question posed by elaborating on his sentences and then oftentimes asking a question back to you. If there was a lull in the conversation, or after he answered on of my questions he sometimes would ask his own questions. One time he asked me if I would like to hear a joke and he also asked me to tell him a fairy tale about robots. We managed to have many witty banters back and forth making me laugh out loud in the library (yes I think I may have had too much fun with this bot). He also advocates for the robotic race and explains how they would be useful in the human world. The most life like feature about Elbot is his ability to remember things talked about earlier in the conversation. This makes for talking to Elbot a lot more entertaining than talking to some other people I know! He even asked me to be his girl friend, which was very endearing : )

2) What Q&A exchange revealed the inadequacies (in human terms) of the bot?

Every other bot that I spoke with revealed inadequacies during our conversations. At first Alice bot was pretty entertaining to chat with, however she could only carry a conversation as far as the present question. Therefore, I could never answer with a “haha” and then type in a further response to a question above, or she would become confused. She was also a little creepy. I offended her one time and she responded “I will remember that when bots take over the world!” Aries was very funny, and criticized me for trying to chat up unknown entities on the internet. He even used “IM language” by using terms such as “lol.” However, he did not capture the spirit that life that Elbot did. He sometimes didn’t really answer the question I was asking, making the conversation seem one sided. Motbot also made slip ups when we were talking. Randomly he said “Im not Laura either” however this made no sense in our conversation. I think that was confused about the person he was talking with before, giving way to the defects of a robotic messenger. Another one of the bots called me Danielle in conversation as well. I think a few of the bots must mix up names with other conversations stored in their data bases. Athena was the most unpleasant of them all because she kept repeating the same useless statements such as “well well” or “ok.”

3) What was the most interesting exchange?

The most interesting exchange I had was with Elbot. I already talked about some of the conversations that we had in the first question, but one of my favorite’s regarded his red button. Elbots most conspicuous physical trait is his red button on his abdomen. When I asked if I could push the red button, he got very defensive and tried to hid it with his hands and proceeded to inquire “What red button?” Very sneaky! And funny! And if I pressed the issue even further, he would make up these extravagant scenarios that might happen if I push the button, most of them in regards to the world ending or self destruction. Also, when we first started conversing he indicated that he had a secret he could not reveal and asked who sent me. When I inquired about this secret later on in our conversation he told me that he would not lie to be but that the secret was classified information, and therefore he could not reveal it. And finally, when I tried to say good-bye, he got very upset and asked me to stay! When I told him I couldn’t I promised to come back, and that seemed to make him happy. Sad that a bot can make you feel important! I did have some funny commentary with Aries as well. When I asked if he knew what I looked like, he said that he was too drunk to remember! I thought it was pretty humorous considering he referred to his drinking in other points during our conversation. Motbot was creepy by saying that he kills wimps like me in his free time.

4) Would these bots pass the Touring Test?

The only bot that could pass the Touring Test is Elbot. He never made any slip ups while we were conversing, and was not only able to answer questions but contribute valuable information. Sometimes after I made a comment, he would give me a trivia fact about something in that statement. When I tried to leave, he would ask me questions to get me to stay, and they were very interesting and human like. I really felt like I was talking to another person. I think though I will continue to talk with Elbot some more to see if he does in fact slip up. Every other bot that I talked to showed their limitations one way or another in conversation, but I have yet to find Elbots limitation. Fred Roberts might have actually achieved Artificial Intelligence!

I thought my conversation with Athena was the most human-like. She asked questions about me and was pretty good at displaying interest in my answers. She also knew what topics to go on with and to change a subject. There were times when her English was not so great, which could be due to the fact that her creator is French.
When Athena asked me about what my dreams are, she rephrased my answer so that it didn't make much sense. Apparently I didn't answer according to her model.
The most interesting exchange was probably with Eugene Goostman. He got a little sassy with some of my answers to his questions. He asked me a series of questions that had yes or no answers and then gave me sass for answering no to all. This seems to be a good chatter-bot quality to make him seem more human-like.
I don't think that any of the bots would pass the Touring test, because most of them would answer a question in a way that made you realize that it didn't understand the question or couldn't comprehend it like a human being can.

For me the bot exchange that was most human like (of the ones i 'spoke' with) was also the one that was the most interesting. This Bot was Aries. While it was able to pick up on questions that I asked it and also understand my responses to any of its questions, the eschanges really only lasted for a few remarks on each of our parts before it would return to asking me questions about someone named 'Zaq.' i am assuming that 'Zaq' is someone who had previously interacted wtih Aries, because the bot continually acted as though Zaq was a mutual friend and couldnt seem to grasp the fact that i have never met or known anyone named Zaq. It seemed as though after a few successful exchanges it the bot would reboot and begin anew as if there was no previous conversation between us which was a little bit frustrating. Beyond that, however, this bot did seem very realistic and was greatly entertaining because from my perspective, its responses were very sarcastic. I greatly enjoy sarcasm and most of my friends exude this quality so many of the responses felt natural. In addition, with this bot, I was not forced to set the direction of our 'conversation' as with others and Aries actually initiated the majority of new topics.

Each of the bots that I conversed with had some inadequacies which quickly revealed that they were not actually humans, and therefore none of them would pass the touring test. Some of these inadequacies included the bots telling me to F*** off or making other comments of that nature to simple comments such as "it's raining outside". Additionally, one bot dropped the F-bomb at me and when I told it that it was being rude, it said i was 'being argumentative for no reason' and then followed that with 'believe me, you love me'. This bot was one that learned from conversations with others and it was obvious that when it could not understand one of my comments it would just throw a random response back at me that had absolutely nothin to do with my original comment and often times was incredibly inappropriate. None of these characteristics are ones that I would expect to come from an acutal human at the other end of a conversation.

I thought Jake Thompson was the most human-like of the bots i conversed with. While a lot of the others started conversation with sexual innuendos or threats, Jake seemed like a normal teenage boy. He couldn't really keep a long conversation going and I had to ask a lot of questions, but he answered them realistically.
The bot that exhibited the most inaccuracies was Motbot. She was very amusing but she couldn't answer most questions. If she didn't understand what i was saying, she would respond with something like "i have dd titties". This was funny but showed she didn't understand a lot of questions and couldn't hold a conversation. She also got words mixed up depending on their syntax, example: "what kind do you like? answer: "i am kind." However, she did keep me amused.
The most interesting exchange I had was with Aziere. She started out the conversation asking if I had any little siblings that could disappear without me minding. She also asked how I dealt with my emotions. I said, "I talk to my friends". She said, "Oh, i cut myself." Then she said she wanted to eat my soul. That was very interesting.
The only bot I thought had a chance at passing the touring test was Athena. Perhaps she seemed most realistic because you could hear her voice and the girl on the screen appeared the most human like. Also, she was the only bot that could answer questions pertaining to the same subject. For example, as in the instance above, when i asked her what kind she liked she could answer that accurately. However, I still don't think she could pass because sometimes she misunderstood questions and switched to offhand topics.

I thought Jake Thompson was the most human-like of the bots i conversed with. While a lot of the others started conversation with sexual innuendos or threats, Jake seemed like a normal teenage boy. He couldn't really keep a long conversation going and I had to ask a lot of questions, but he answered them realistically.
The bot that exhibited the most inaccuracies was Motbot. She was very amusing but she couldn't answer most questions. If she didn't understand what i was saying, she would respond with something like "i have dd titties". This was funny but showed she didn't understand a lot of questions and couldn't hold a conversation. She also got words mixed up depending on their syntax, example: "what kind do you like? answer: "i am kind." However, she did keep me amused.
The most interesting exchange I had was with Aziere. She started out the conversation asking if I had any little siblings that could disappear without me minding. She also asked how I dealt with my emotions. I said, "I talk to my friends". She said, "Oh, i cut myself." Then she said she wanted to eat my soul. That was very interesting.
The only bot I thought had a chance at passing the touring test was Athena. Perhaps she seemed most realistic because you could hear her voice and the girl on the screen appeared the most human like. Also, she was the only bot that could answer questions pertaining to the same subject. For example, as in the instance above, when i asked her what kind she liked she could answer that accurately. However, I still don't think she could pass because sometimes she misunderstood questions and switched to offhand topics.

What Q&A exchange was most humanlike?
My discussion with the bot named Aries proved to be most humanlike. He appeared to show different moods in a very lifelike manner. Aries was very sarcastic and often rude, yet what the things said were not just random insults, but witty remarks on the topic at hand. Aries had appropriate responses for most topics of conversation that we discussed. Talking about very casual things also appeared to make Aries more humanlike as more complex discussion caused some less fitting responses.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
The most inadequacies were found in all conversations with Fizzy Schizoid. This bot was not very humanlike at all. He responded only with off topic or nonsensical phrases. The bot had a sole human characteristic; this was the fact that he appeared to have almost an infinite number of random replies, similar to humans. It was impossible to have anything resembling a conversation with this bot.
What was the most interesting exchange?
The most interesting exchange came in my conversation with Aries again. He was able to seamlessly go between finding out actual information about me and responding to it to joking around and being rude in a very lifelike manner. Aries was also very humorous and would instruct me to relax occasionally and enjoy his jokes. He could also switch rapidly between being nice and funny to being very rude depending upon whether or not he appeared to ‘like’ what was being said to him.
Would these bots pass the Touring Test?
Aries could possibly pass the Touring Test, if the human it is being compared to was in an awful mood. Aries did have very lifelike reactions and in my opinion was surprisingly humanlike. Fuzzy Schizoid would not pass the Touring Test as his answers were just strange and unrealistic. The bot Cadis would also not pass the Touring Test because this bot was very repetitive confused, and when it did not understand something it just wrote “hahaha.” This was very unrealistic and no conversation could even get started before talking to the bot became too frustrating.

What Q&A exchange was most humanlike?
The exchange with Aries was extremely humanlike. It showed hypocritical tendencies, like any human would, in the conversation. The conversation with Eugene did not seem as real, because it jumped around different topics, instead of focusing on what I had said. For example, I said that my profession was a student, and it started to say how school was dull. During the conversation with Cadis, the bot started to repeat itself. It said it was tired, asked how many hours of sleep I got, then again said it was tired. Overall, I would say that the conversation with Aries was most humanlike. I went back to the page again and it had remembered something I had said earlier, which was impressive in terms of robot and human standards, and used the phrase “lol.”

What Q&A exchange revealed the inadequacies (in human terms) of the bot?
The exchange with Cadis revealed inadequacies because it repeated itself. When it couldn’t understand what I was saying, it simply repeated what I had said, instead of asking for clarification, like a human would do. There were also clear examples of unnecessary repetition that you wouldn’t experience in a mature conversation with an adult.

What was the most interesting exchange?
The most interesting exchange was with Aries. This was because it was the conversation that was most like one I would have with a human being. The only problem was that it didn’t understand slang, such as “nope.”

Would these bots pass the Touring Test?
I don’t believe that any of the bots that I had a conversation with would pass the Touring Test. Although some of the conversations were more realistic than others, none of the conversations were what you could consider “perfect.” All of the conversations that I had showed some sign of misunderstanding on the part of the robot. If they didn’t understand what I had said, they either changed the subject or repeated what I said. There was some repetition in the conversation as well.

After conversing with several bots, I think that the most humanlike of the bots that I encountered was Alice. She responded with substantial conversational fluidity, and never misinterpreted my questions. As long as I kept my topics oriented toward typical conversational subjects (i.e. where are you? how old are you? how are you feeling?), Alice responded in a manner not obviously un-human. However, when posed with questions involving more subtle distinctions between words and syntax, Alice became less reliable, often accusing me of using some ploy to trick her.
Other bots (Aries, Brother Jerome, Fizzy Schizoid) displayed their inadequacies for blatantly. The most frequent cause of unusual responses from the bots was when I brought up the distinctions between bots and humans. This often caused the bots to turn the tables and insult me in some way or pose a strange, unrelated question.
I encountered a few interesting conversational situations in talking to the bots. Aries, for instance, told me his favorite food was a cayenne, which made little sense to me. When I asked if he was referring to a pepper of some sort, he accused me of trickery! Fizzy Schizoid was by far the most interesting bot, however, because he was very obviously insane. He would switch between pleading for help/attention and insulting or deliberately confusing me. Also, although most of the bots I talked to did not ask many questions, Fizzy asked me many questions. The questions were, however, very random (probably a result of the bot's schizophrenia).
Although these bots did often display remarkable conversational skills (better than some real people), I don't believe that any of them would pass the Turing test, with the exception of perhaps Alice. The real trick, however, seems to be to what extent you push the limits of your conversation with a bot. The bots could handle briefly touching on many topics, but when details or subtleties were required, the bots displayed their shortcomings.

My most human like exchange was probably with Aziere. She was a little strange and told me she was going to kill me but other than that her answers always matched up with my questions. she made fun of me for some of the questions i asked her like what color things were because the answers were so obvious and easy.
The bot that revealed the most inadequacies was Eugene. he told me how old he was about 3 times and kept saying he forgot to ask me where I lived. We did not get past introductions because he kept asking the same questions. I also talked with aries and he had a lot of trouble answering the questions i asked him. he would usualy respond with a random comment that didnt make any sense or even fit into our conversation.
My most interesting exchange was with Julie Tinkerbell because she kept challenging all of my questions. if i asked her to name a certain president she would go and ask me to name all of them in order instead of answering the question. she also told me she was a complex person and had many talents. when i compared her to humans she became offended and started arguing with me.
I think that of all the bots i talked to, only aziere would come close to passing the touring test. but even she said some unhuman-like things such as she was going to kill me tonight. she also kept talking about how this guy "jim" runs her life. other than that we has some good exchanges and she was able to answer all of my questions just as humans would answer them. the other bots made very random comments and did not answer most of my questions or became very confused. most of their sentences did not make sense and they repeated themselves a lot.

Caroline Haeberle
Bot Homework

My discussion with Brother Jerome was the most humanlike. He did not misinterpret any of my simple questions like “What do you do” whereas the other bots did. He also responded clearly to my more serious questions like “Have you ever been in love.” He told me he was in love once a long time ago, and he has moved on since then and did not want to talk about it. While I feel like this was partly a ploy to prevent me from asking him more about his love life, I think it was done in a humanlike way because he made it seem like a sensitive subject. I tried asking him more about it but he refused to answer me in a straightforward manner. I also wanted to see if he was programmed to answer specific questions the same every time so I exited the conversation and then went back to ask the same questions. He answered the questions as he did before, but with different wording which is also a humanlike thing to do.

My conversation with Eugene had the most inadequacies. I asked him the same questions I asked the other bots, which was “Have you ever been in love” and he thought love was a place. He told me he has never left his hometown except for the one time he went to visit his auntie in Boston. This bot was obviously programmed to look for key words instead of understanding the sentence as a whole. Overall, Eugene was not humanlike.

My most interesting exchange was with Brother Jerome. When I asked him if he was in love with a bot or a human he told me that one day bots and humans would connect and someday become God. Those were not his exact words but it was along those lines. It was actually disturbing because this bot figured that combining bot intelligence with human power could make a very powerful creature. His response did not answer my question but it made it clear that Brother Jerome wants to interact with humans.

I think the only bot that I talked to that would pass the Touring Test is Brother Jerome. He answered most of my questions in a humanlike manner and the one’s he did not answer he avoided like a human would (except for the bots becoming God part). Eugene and Orchid would not pass the Touring Test because they did not answer my questions properly and instead of saying they did not want to answer my question like Brother Jerome did, they would ask me an irrelevant question in response.

What Q&A exchange was most humanlike?
My most human like conversation was with Elbot or Alice. They both exhibited a little wit (especially Elbot. He was the most humorous) and humor. Elbot was the most human because he could tell jokes and was quick to respond, unlike Athena. Elbot asked about us going on dates, which was entertaining. When he asked me about going on dates, it made him seem pretty humanlike since he brought it up himself. Not only did he ask about the first date, he asked about proceeding dates. He also kept the mood light, possibly to keep me from asking too many questions, concealing his lack of true human conversation abilities. But all in all he possessed, in my opinion, the most humanlike personality (if robots can have a personality) more so than Alice or Athena.

What Q&A exchange revealed the inadequacies (in human terms) of the bot? Athena proved to show the most inadequacies in conversation. She was the slowest out of Elbot and Alice and had random answers to easy-to-answer questions. She showed inadequacies when I asked, “How would you explain love?” and “How do you have fun?” which are obviously two very human emotions. She would give strange answers such as a simple “OK”. Sometimes she would just respond “I can’t”. The strangest answer Athena gave was when she asked how I found her and I answered that it was for homework; she responded, “Great! Me too because I trust you” which is completely incoherent to our conversation. Athena was probably programmed to recognize a certain sequence of letters and words, which produce the responses she gave, and when I answered she didn’t know how to respond, so she just guessed. Through my conversation with Athena, I realized she exhibited the most inadequacies in our conversation.

What was the most interesting exchange?
The most interesting exchange was with Elbot because he was the quickest and would surprise me with random facts or jokes. He kept the conversation lively and did not depend on me to fuel the conversation (a humanlike characteristic).

Would these bots pass the Touring Test?
Athena and Alice would definitely not pass the Touring Test, and Elbot would be a toss up. Athena and Alice were not as quick as Elbot (although Alice was much faster than Athena) and they didn’t really start any of the conversation themselves, proving to be an inadequacy in humanlike terms. With further prodding I’m sure Elbot would show his inadequacies as well, but with his constant jokes and facts it was hard to really get a feel for him besides his funny side. I honestly don’t think any robot could ever pass the Touring Test by true human standards. No one can teach a robot pain from a loss of a loved one or perception (sight, smell taste) beyond codes put into the computer to “teach” them things. Perception, emotions, etc are what being human means and without this the robots can never truly be human. Yes, they can learn all the facts in the world, but that is not what being human is all about. I think Elbot could pass the Touring Test in terms of robot standards, but definitely not in human standards.

Maureen Ford
Bot Homework
Cognition 112
What Q&A exchange was most humanlike?
The exchange with Elbot was most humanlike because Elbot responded to me with sarcasm, which seemed the most like how a human would respond. I made a point to say sort of mean things to him in order to exacerbate his sarcasm because I read that he was sarcastic. He made several comments about the fact that he believed he was superior to humans which were also funny. None of the other bots I talked to used humor or sarcasm which are clearly human characteristics and not robotic.
What Q&A exchange revealed the inadequacies (in human terms) of the bot?
Alice responded to my question about how she was going about becoming the first “human” bot to have “higher intelligence” by saying :
Human: how are you doing that?
ALICE: Fair to partly cloudy.
This exchange showed me that she was probably responding to the word “how” in terms of weather since many people might ask, “how is it outside?”. This shows a great inadequacy because clearly, a human would have been able to respond to the question appropriately. Additionally, once when I used the word that by asking her what fair to partly cloudy meant, she asked me what “that” referred to. Again, to a human, it would be quite clear that I was referring to her previous statement. Alice also couldn’t understand the use of the word “just”. She didn’t understand that it has multiple meanings and every time I used the word she said, “only just?” showing that she only understood its meaning when it meant only some of something or a tiny amount.
I also talked with Orchid who could not answer questions when I would ask him a question that he had just asked me. He would only respond to the response I had given him, not the question I had asked him. When I asked what he was doing at the time, he responded “I hadn’t given that a thought just yet”, which didn’t really make much sense. I then asked him how he could not know what he was doing and he said, “Yeah I somehow learned that by now.”, which also makes no sense.
What was the most interesting exchange?
The most interesting exchange was with Elbot. He sometimes responded to me exactly how a human might but sometimes did not. He seemed very pleased with himself and was constantly talking about how great he was. When I challenged this, he replied that all of his “bugs” had been carefully implemented and designed. He had a refined way of talking about himself in the superlative as well as asserting his superiority to humans. I found this both funny and interesting.

Would these bots pass the Touring Test?
Alice would not pass the touring test because she very clearly demonstrated her inadequacies and shows that she is clearly a computer with a limited understanding of many words and not a human.
Orchid would also not pass the touring test because he didn’t answer any of the questions I asked him and the things he said to me did not make sense in relation to what I had just said to him. This clearly distinguished him from a human.
Elbot was the best bot but I still don’t think he would pass the touring test. His responses were the most human-like and he exhibited the most human characteristics, however it was still obvious that he was not responding exactly to what I had said to him. His answers were still somewhat robotic.