It is 6 30 pm here in Redmond Washington but on  the other side of the world in shanghai it is like  

10 30 a.m and so i'm going to be chatting  with Zihao who is a Microsoft student  

ambassador and an mvp and he has an insane  video channel in china where he's teaching  

like hundreds of thousands of people how to code  and how to get started in artificial intelligence  

and the community that he's managed to  build up in a really short period of time  

is just amazing to me and so I had to sit  down with him to hear about what he's doing  

how we got into coding and what his code  story is so let's uh let's check it out  

hey how are you yeah hi thanks for this  opportunity awesome fantastic well tell  

us a little bit about yourself um hi my name is  Zihao a student from tongji university shanghai  

and i have a video channel about artificial  intelligence machine learning python programming  

deep learning computer vision and raspberry pi  and have many tutorials of Azure and cognitive  

service such as custom vision I'm looking forward  to sharing my experience about technology open  

source content creation and community. I love  it, I love it. So we were talking a little bit  

earlier and you were telling me about how you  got into computer science and programming and  

it's a little bit different than what you were  planning on doing when you went to college so  

tell me a little bit about how you switched your  focus areas i majored in civil engineering when  

i was an undergraduate to be honest at that  time i hate coding i never write a single  

line of code my major civil engineering focuses  on all kinds of construction and infrastructure  

such as design and build architecture house tunnel  bridge and the railway nothing but mechanics and  

dynamics the thing changed in 2018 when i was an  exchange student in the university of Pennsylvania  

there was a lecture in Wharton school inviting  the CEO of Microsoft Satya Nadella he shared his  

insights about blockchain quantum computing as  well as open source communities such as GitHub  

most people sort of say it's the ai versus the  human i think we have to sort of first grab hold  

of all the opportunities ai creates to empower  humans later i found GitHub this amazing website  

i learned lots of open source toolkits such as vs  code keras even bitcoin on this website in GitHub  

you can collaborate with developers all over the  world and i started to study python programming  

data analysis algorithms machine learning ai deep  learning computer vision front-end development  

with eager and hunger of course that was not  an easy way i used to get stuck in a python bug  

or dealing with data structure issues for the  whole afternoon or the whole day at the same time  

i started to write technical blogs and make  technical videos and and so you were making  

these videos and writing these blogs because you  didn't want other people to have to maybe struggle  

the way that you were struggling in the beginning  right my first video was about convolutional  

neural networks let's break through algorithms  in deep learning and the computer vision  

help us to do image classification objective  detection and image recognition later i found  

myself good at making this kind of hardcore  tutorials and the more and more subscribers  

make me more confident and i think this could  be the career of my lifetime we aim to help  

zero beginner and the cross-field learners like  me quickly master data mining machine learning ai  

in easy to understand the way and quickly  apply what they have learned into real scenario  

projects i love that actually my video  channel covers all kinds of fields related to  

computer science especially in computer vision my  video channel covers mathematics machine learning  

basics deep learning neural networks computer  vision python programming data analysis data  

visualization and edge computing especially we  focus on computer vision because computer vision  

is telling a computer how to extract information  or recognize information from images or videos  

and the computer vision also called cv is a  basic application in artificial intelligence  

it contains image classification object detection  semantic segmentation instance segmentation and  

also object tracking 3d vision let me show  you an example okay this is object detection  

object detection means drawing rectangle boxes on  an image and with this algorithm you can do fine  

tuning and transfer learning if you are a doctor  you can use this algorithm to detect tumors and  

pneumonia on x-rays and if you are a farmer you  can use this algorithm to detect a crop disease  

that depends on you this can empower lots of  fields and a lot of professions and is widely used  

in different scenarios and a lot of students from  traditional majors they can really benefit from  

computer vision that's such a good you know demo  and a good visualization of what you're doing what  

i guess kind of led you to continuing to create  these videos were you making them as you were  

as you were continuing to learn exactly more and  more people in china like me especially students  

from those traditional majors such as mechanical  finance they want to get into the it world too  

they want to use advanced machine learning  to empower their own profession and their  

own major i think with my experience i believe i  can really help them simplify the learning curve  

let me show you another example i major in civil  engineering and especially high-speed railway  

and the chinese have a huge high-speed  railway and urban railway system and how  

to detect and maintain the health of this kind of  infrastructure is of vital importance that's what  

i can do to break the ceiling of the traditional  engineering this video is detected railway track  

shelling tracking checking damage corrugation and  the fatigue fracture on the running surface of the  

railway and we can expect that in the future  everything is related to data and everything  

is data-driven yeah no we were talking about that  earlier about how important you think it is that  

people actually try things in order to learn them  and that that's how you've learned a lot is by  

by doing things can you talk a little bit about  that for example if you are an agriculture expert  

or a farmer this is a traditional job with a long  history you can utilize new technologies such as  

uav drones remote sensing and computer vision  to do lots of things breaking the selling of  

traditional agriculture now you don't even need  to write code or learn algorithms there are lots  

of out of box toolkit such as custom vision in  azure cognitive service you can just upload your  

data sets train and deploy the ai models just by  clicking your mouse this is also a demo to do the  

crop disease image recognition if you are a farmer  or an expert in agriculture you just need to  

upload your photos and the artificial intelligence  will detect and recognize which disease does  

your crop have with high accuracy and the high  precision i love that and you have you have 340  

000 subscribers across different platforms is that  right yeah and in the future there will be more  

which is incredible but and you  started doing this in what 2018  

yeah i started from 2018 but during the pandemic  the online education became more and more popular  

that's how i get more popular it would make sense  that students especially would really get value  

out of your channel and especially the fact that  you know you present things in an interesting way  

and you use real world examples so it makes total  sense that this is where a lot of your subscribers  

have come from also like just like i said before  you don't need to write code or even you don't  

need to learn algorithm because there are already  many out of the box toolkit such as custom vision  

in every cognitive service you just upload your  data and the error will help you handle everything  

help you train or deploy the models you just need  to click the mouse curious we were talking earlier  

and you said that there haven't been a lot of  resources for people in china who speak mandarin  

to you know learn we have a lot of documentation  in english but not as much in in mandarin so  

has that been a good opportunity for you  i have made a series of videos to help you  

read academic paper of artificial intelligence  especially computer vision i took detailed notes  

of every formula every graph every reference the  paper and what's so interesting you know you're  

showing these papers that you've annotated in  chinese you've added these notes but you were  

telling me that some of your content that you  create in in chinese has actually been translated  

into other languages for for other people in other  parts of the world have you had a lot of people in  

your community helping contribute to your content  as well yeah actually i have a whole i have a big  

team the team member not only from my own team  my own school but also from my subscribers in all  

kinds of fields and all kinds of professions  i love that you're focusing on all those  

ways to to you know get the most out of what  you're doing and then also helping people  

in all parts of the world potentially  be able to do that too it's not just  

one thing which is what makes open source so  great is that it can be global and it's what  

makes the community that you're building so great  so in addition to having your your video channels  

and your tutorials i know that you also work as a  an mvp and a microsoft student ambassador can you  

tell me a little bit about the work that you do  there microsoft supports microsoft student club  

in hundreds of universities in china also in  my college tonji university shanghai we have  

the opportunity to work with microsoft experts  from microsoft research asia also called msra  

on various cutting-edge projects we make a whole  open source community of ai education called aiedu  

let me show you a github page that's  awesome and also there are many chinese  

college students take part in this open source  repository and contribute themselves and mvp  

means most valuable professional and mvp is the  technical expert and community leader as well as  

open source advocates not only do they write blogs  and make videos but also they will give feedback  

on microsoft products and we also take part in  important activities in the community such as  

microsoft ignite conference a microsoft build  connect conference ai bootcamp and i'm msa means  

microsoft student ambassador and msa members come  from university students all over the world many  

of them are college students you can make friends  with them and you can communicate technical issues  

and even report a bug that's awesome have you made  a lot of friends um in in the community yeah and  

msa members from all over the world i love  that i love that i i have a question for  

you for if somebody is whether they're a  student or maybe they're just looking to  

get started what's your advice for someone who  is coming from a different background you know  

like you who wasn't a a typical computer science  major what's your advice for them on on getting  

started here is the learning path i recommend you  need to take four crucial steps and the first is  

mathematics such as linear algebra probability  statistics and calculus and the second step is  

coding and the programming i recommend python and  javascript and python is popular not only because  

it's it's a very easy to learn programming  language but also they have python have a huge  

such as csikit learn and you can use them to  make your own application so learning python  

is of vital importance in this period of time  and the third step is algorithm if you want to  

learn more about artificial intelligence and  deep learning you need to know the basics of  

convolutional neural networks recurrent neural  net networks uh all kinds of neural networks  

you need to learn the algorithms basic and the  last and the most important step is a real project  

which means get your hands dirty you need to learn  through the real scenario project dealing with the  

real scenario a problem so what's next for you  you know you've been you've been making videos  

um a lot during uh you know the the pandemic and  um you've been learning a lot what's what's the  

next thing that you want to focus on yeah i have  a huge plan i will build my ai education ecosystem  

in two years that's my dream this process is  extremely enjoyable for us from the bottom  

of our hearts that we are providing high  quality education resources for a huge number of  

future learners i love that i love that i love  that you're helping people all over the world  

yeah well thank you so much for talking with me  and for doing all the great stuff that you do  

this is so cool and and i i love um just seeing  someone like you who's so driven and who's so  

um willing to want to teach other people and i  think that you've found a really good career and  

and i think that this is something you're meant  to do i'm so happy to have talked to you yeah  

maybe two years later we can have another code  story interview let's do it let's uh let's plan  

to talk in two years because i can't wait to see  what you um what you build then i'm so impressed  

with everything you're doing uh i think it's so  great two years from now we're definitely going to  

talk again because i can't wait to see what more  you've accomplished thank you so much thank you  

you know just a couple of years ago Zihao was  a civil engineering major and he didn't have  

any interest in programming at all he saw satya  nadella microsoft ceo speak when he was at the  

university of pennsylvania and it opened up this  whole new world for him and not only has he become  

a really proficient programmer but he's actually  gone on to teach you know thousands of people  

from all over the world how to get started with  artificial intelligence and machine learning and  

python and open source and he's building out an  entire curriculum to really empower people i think  

that's really incredible and it just goes to show  you that all of our code stories can be really  

unique and can come from really different places  and the fact that so many more people are going  

to be learning because of him is really exciting  we've got some resources down below where you can  

check out how to get started if you're interested  in learning both at docs.microsoft.com and  

microsoft learn i also have some links to Zihao's  github page because all the ai education work that  

he was doing while an intern at microsoft research  is just incredible and you know i can't wait to  

see what he does next we promised that we would  talk in two years so we will but i'm going to head  

out now because it's a little bit late here but  we'll see you next time thanks for tuning in you