My Journey to Artificial Intelligence
Andrew Widjaja, Technical Leader IBM Software — Sinergi Wahana Gemilang
Artificial Intelligence or AI, these words are familiar to those already playing in the computer world.
I started to be interested in AI since I was in college and watched several science fiction movies, for example, Chappie (2015). In that movie, there’s an interesting story. The base-components of a robot are just a steel-frame (or metal skeleton) and a few sensors. However, how do we make the robot so it can make decisions based-on experience? Of course, conventional programming cannot solve this problem, such as traditional software development in general.
My knowledge at that time about AI was minimal. I felt that the AI course at that time was challenging. In that course, I learned several types, methods, and basics of AI, such as Linear Regression, Classification, Clustering, and others.
My learning about AI basics continues in my final project entitled Detection of Vehicle’s Speed on the Road by using video as input data and testing data. For vehicle object recognition, I used the computer vision library, OpenCV (open source). Initially, I wanted to train the data manually but unable to do so because of supporting infrastructure constraints.
In the end, I did vehicle detection using blob detection and noise reduction. After the vehicle detection is successful, I analyzed the vehicle’s speed by installing 2 lines on the video, the start line, and the finish line, and seeing the travel time from both lines.
I got a deeper knowledge of AI when I started my career at PT. Sinergi Wahana Gemilang (SWG), IBM Value Added Distributor, in 2017. My title at that time was Solution Specialist for Emerging Technologies. My colleagues and I have made several AI showcases, such as Image Captioning with IoT devices and a combination of Face Recognition with IoT devices, for IBM events.
In SWG, besides studying AI, I also studied Data Analytics because the two areas intersect with each other. I have been learning about these 2 areas through forums/classes, such as AI Summer School, and online learning, such as Cognitive AI class (https://courses.cognitiveclass.ai/) and Data Quest (dataquest.io).
AI Summer School, which consists of several classes, is held annually with several organizers and sponsors, such as the National University of Singapore and the Nvidia Institute. I attended AI Summer School in 2019, Singapore. Most of the speakers were experienced lecturers (professors), and most of the attendees were Singaporean (students).
I often follow the Cognitive AI class by IBM. In online learning, I took several courses and badges related to Data & AI, starting from how to build chatbots, data science methodology, and text analytics. On the coding side, I followed the Data Quest to learn how to process data using Python.
For me, both are the best online learning. In Cognitive AI class, we were presented with videos from the experts and some quizzes. In contrast, in the Data Quest, we were presented with a problem, and we had to solve the problem using the Python programming language.
From my experience so far, if we want to master AI truly, we must understand the math behind machine learning, comfortable with abstract thinking, learn how to gather the right data, and most importantly, have the passion for exploring, such as searching the internet for different types of AI, online courses and join online communities.
