Artificial Intelligence has grown tremendously in the past decade. AI-enabled tools have brought about remarkable changes in our society. Powerful machine learning programs have been trained, deployed, and applied in various fields.
As AI/ML has become much more popular with the mass audience, it’s time to build the public’s trust in AI by developing successful AI systems through smartsizing data and having them help users gain new knowledge, increase productivity, and cut unnecessary costs.
This is where data-centric AI and explainable AI come into play.
In a recent interview with Dr. Phong Nguyen from FPT Software, Wow AI had the chance to explore the applications and importance of data-centric AI and explainable AI in the biggest software services company in Vietnam.
Watch the whole interview here
About the speaker
Phong Nguyen is currently the Chief Artificial Intelligence Officer at FPT Software. He is responsible for developing AI capabilities and strategy for the organization and building the Research and Development Lab, as well as the Solutions Department to provide state-of-the-art AI. Before joining FPT, he was a researcher at Hitachi Research Lab and a visiting researcher in MILA (Montreal Institute for Learning Algorithms) after obtaining a Master's degree from Carnegie Mellon University and a Ph.D. from the University of Tokyo in Japan. MILA is one of the top AI Institute research labs in the world, led by Professor Yoshua Bengio, a Turing Award winner for his contribution to AI.
In the upcoming Worldwide AI Webinar 2022, Phong Nguyen will be discussing the data-centric AI approach and the future of practical AI.
On data-centric AI
Phong Nguyen is a true advocate for data-centric AI and its practicality. He shared that though AI and ML are essential software for businesses and enterprises, not many organizations can harvest their power.
To build practical AI, enterprises need to take a different perspective from academia, focusing on data-centric AI instead of model-centric AI. Data-centric AI can help companies to overcome challenges such as limited data, noisy data, and low-quality data.
Why explainable AI (XAI) is important
According to Phong Nguyen, XAI helps users understand and trust the decisions of the AI model, or in his words “unbox black box AI”. Since some AI models are too complex for users to comprehend, explainable AI comes in handy to enhance the customers’ trust in the machines.
He shared that FPT Software is currently using explainable AI to understand the reasons and properties leading to high or low happiness levels among their employees and improve accordingly to retain talents longer. Putting XAI at the core of the business and making it one of their three pillars besides data-centric AI and AI for software, FPT Software is enabling customers’ belief in AI and helping improve transparency as well as productivity.
Watch the conversation between Phong Nguyen and Wow AI here