{"title":"What Is AI?","authors":"I. Millington","doi":"10.1201/9781003124047-2","DOIUrl":null,"url":null,"abstract":"Siri is not the only virtual assistant that will struggle to answer this question (see Fig. 2.1). Toma et al. (2008) showed that almost two thirds of people provide inaccurate information about their weight on dating profiles. Ignoring, for a moment, what motivates people to lie about their dating profiles, why is it so difficult, if not impossible, for digital assistants to answer this question? To better understand this challenge it is necessary to look behind the scene and to see how this question is processed by Siri. First, the phone’s microphone needs to translate the changes in air pressure (sounds) into a digital signal that can then be stored as data in the memory of the phone. Next, this data needs to be sent through the internet to a powerful computer in the cloud. This computer then tries to classify the sounds recorded into written words. Afterwards, an artificial intelligence (AI) system needs to extract the meaning of this combination of words. Notice that it even needs to be able to pick the right meaning for the homophone “lie”. Chris does not want to lie down on his dating profile, he is wondering if he should put inaccurate information on it. While the above steps are difficult and utilise several existing AI techniques, the next step is one of the hardest. Assuming Siri fully understands the meaning of Chris’s question, what advice should Siri give? To give the correct advice, it would need to know what a person’s weight means and how the term relates to their attractiveness. Siri needs to know that the success of dating depends heavily on both","PeriodicalId":141388,"journal":{"name":"AI for Games","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"AI for Games","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1201/9781003124047-2","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Siri is not the only virtual assistant that will struggle to answer this question (see Fig. 2.1). Toma et al. (2008) showed that almost two thirds of people provide inaccurate information about their weight on dating profiles. Ignoring, for a moment, what motivates people to lie about their dating profiles, why is it so difficult, if not impossible, for digital assistants to answer this question? To better understand this challenge it is necessary to look behind the scene and to see how this question is processed by Siri. First, the phone’s microphone needs to translate the changes in air pressure (sounds) into a digital signal that can then be stored as data in the memory of the phone. Next, this data needs to be sent through the internet to a powerful computer in the cloud. This computer then tries to classify the sounds recorded into written words. Afterwards, an artificial intelligence (AI) system needs to extract the meaning of this combination of words. Notice that it even needs to be able to pick the right meaning for the homophone “lie”. Chris does not want to lie down on his dating profile, he is wondering if he should put inaccurate information on it. While the above steps are difficult and utilise several existing AI techniques, the next step is one of the hardest. Assuming Siri fully understands the meaning of Chris’s question, what advice should Siri give? To give the correct advice, it would need to know what a person’s weight means and how the term relates to their attractiveness. Siri needs to know that the success of dating depends heavily on both