Given the task definition and input, reply with output. In this task, you're given a question, along with three passages, 1, 2, and 3. Your job is to determine which passage can be used to answer the question by searching for further information using terms from the passage. Indicate your choice as 1, 2, or 3.

Question: How old is the person who succeeded Hastings? Passage 1:Kanazawa was born in Iruma on July 9, 1976. After graduating from Kokushikan University, he joined J1 League club Júbilo Iwata in 1999. Although he could not become a regular player, he played many matches as left side midfielder from first season. The club won the champions 1999 and 2002 J1 League. In Asia, the club won the champions 1998–99 Asian Club Championship and the 2nd place 1999–00 and 2000–01 Asian Club Championship. In 2003, he moved to FC Tokyo. He became a regular player as left side back from first season. The club won the champions 2004 J.League Cup. Although he could hardly play in the match for injury in 2006, he came back and became a regular player again in 2007. From 2008, he lost regular position behind newcomer Yuto Nagatomo and he also played as defensive midfielder not only left side back. In August 2009, he moved to Júbilo Iwata for the first time in 7 years. He played as regular left side back in 2009 season. Although he could not play many matches from 2010, the club won the champions 2010 J.League Cup. His opportunity to play decreased from 2011 and he moved to J2 League club Thespakusatsu Gunma in 2014. He retired end of 2014 season at the age of 38.
 Passage 2:Born in Scarborough, Ontario, Canada in 1957, he attended Lord Roberts Public School, graduated from Midland Avenue Collegiate Institute, holds a BA from the University of Trinity College, University of Toronto, law degrees from Osgoode Hall Law School and the London School of Economics, and was a practising barrister. He moved to New Zealand in 1985. Before becoming Chief Censor, he was Deputy and Acting Chief Censor from December 1998 to October 1999, Senior Lecturer in Law (teaching Legal System and International Law), Deputy Dean of Law, and a member of the governing Council, at Victoria University of Wellington. He was also briefly the Video Recordings Authority in 1994, a member of the Indecent Publications Tribunal from 1990 to 1994 and Deputy President of the Film and Literature Board of Review from 1995 to 1998. In 2010 he stood down as Chief Censor when he became a District Court Judge and Chair of the Immigration and Protection Tribunal. He was succeeded by Andrew Jack.
 Passage 3:Most Deep Learning systems rely on training and verification data that is generated and/or annotated by humans. It has been argued in media philosophy that not only low-payed clickwork (e.g. on Amazon Mechanical Turk) is regularly deployed for this purpose, but also implicit forms of human microwork that are often not recognized as such. The philosopher Rainer Mühlhoff distinguishes five types of "machinic capture" of human microwork to generate training data: (1) gamification (the embedding of annotation or computation tasks in the flow of a game), (2) "trapping and tracking" (e.g. CAPTCHAs for image recognition or click-tracking on Google search results pages), (3) exploitation of social motivations (e.g. tagging faces on Facebook to obtain labeled facial images), (4) information mining (e.g. by leveraging quantified-self devices such as activity trackers) and (5) clickwork. Mühlhoff argues that in most commercial end-user applications of Deep Learning such as Facebook's face recognition system, the need for training data does not stop once an ANN is trained. Rather, there is a continued demand for human-generated verification data to constantly calibrate and update the ANN. For this purpose Facebook introduced the feature that once a user is automatically recognized in an image, they receive a notification. They can choose whether of not they like to be publicly labeled on the image, or tell Facebook that it is not them in the picture. This user interface is a mechanism to generate "a constant stream of  verification data" to further train the network in real-time. As Mühlhoff argues, involvement of human users to generate training and verification data is so typical for most commercial end-user applications of Deep Learning that such systems may be referred to as "human-aided artificial intelligence".
2