I Think a A.I is a computer program that can respond in different situations and can not trick a human into thinking it is a human because current A.I is less flexible than a human and much more formal

What is A.I?

Sources: Tech Target and IBM https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence

https://www.ibm.com/topics/artificial-intelligence 

A.I is short for artificial intelligence. A.I is simply a description of technology such as machine learning that can be intelligent. A.i can analyze data and chatbots like the ones we are testing learn how to respond to text messages. A.I focuses on cognitive skills such as learning, reasoning, self correction, and creativity. However, A.I has many disadvantages such as it is expensive, it needs lots of technical maintenance, and lacks the ability to generalize one task to another. There are 2 types of A.i, weak and strong. Weak Ai can only do one task and is the kind we are experimenting with.

Uses of A.I

Sources:IBM link:https://www.ibm.com/topics/artificial-intelligence

A.I has many applications that we use all the time. It can recognize voices and is useful in voice assistants like Alexa because they need to understand what you are saying. A.I is also used in computer vision where computers/A.I can recognize visual information. This is used in self-driving cars and photo-tagging on social media. 

Manipulated variable: The person or A.I responding

Controlled variable: The starting prompts and voters

Responding variable: what the voters think

Two 4 of paper with one conversation of a.i and human, a.i and a.i, and another one with human to human. We have a sign up sheet for people (voters) to guess which conversation is which (all prompts are refreshed 4 times and the best is chosen)

In test 2 the people seemed more uncertain after about 3 people went and people started to guess incorrectly. Also in test 1 more people were prepared because they knew beforehand but in test 2 barely anybody showed up so we took people at the last minute making them unaware of the challenge beforehand.  In general grade 5 did the worst and grade 4 did the best. Because of the small age difference, this is probably coincidental.                                                                               

In test 1 where voters could see each other's votes, Everybody got 100%. On test 2 where people could not see each other's votes, only 42.8% of people got all correct. Only 1, 6th grader got them all correct and 2, 5th graders. Many people mixed up the Human and Human and Human and A.I which was the most common mistake. Overall, 63.6% of people got them all correct.11 people showed up 7 were grade 5 2 of them were grade 4, and 3 were grade 6(2 grade 4 and 2 grade five were in the first test).

Conclusion

The A.I was mildly successful in tricking the testers into thinking it was a human and made most of the voters think when guessing which converstion is witch and our hypotheses was not exactly correct.

Uses of A.I

Sources:IBM link:https://www.ibm.com/topics/artificial-intelligence

A.I has many applications that we use all the time. It can recognize voices and is useful in voice assistants like Alexa because they need to understand what you are saying. A.I is also used in computer vision where computers/A.I can recognize visual information. This is used in self-driving cars and photo-tagging on social media. 

Sources of error:

All items from A.i can have some edits and not all things that are given by A.I will be included to prevent obvious giveaways for voters. Also some prompts (not the first one) may be different, but we tried to stick to a common theme.

Thank you to IBM and tech target for helping us with research and our parents and teachers and parents for guiding us through with our subjects for data including Catherine, Kate, Ben, Ayaan, Nic, Zander, Austin, Bharath, Ethan, Kaylin Mustafa, and Radien. Thanks to all these people, as we could not do this without you!

Credits

Thank you to IBM and tech target for helping us with research and our parents and teachers and parents for guiding us through with our subjects for data including Catherine, Kate, Ben, Ayaan, Nic, Zander, Austin, Bharath, Ethan, Kaylin Mustafa, and Radien. Thanks to all these people, as we could not do this without you!