Visiting BAAI (Beijing Academy of AI)

vbox10031__18A0973_141459_small.JPGI was invited to give a talk at the Beijing Academic of AI conference (BAAI). I presented our work on the relation between product relevance and customer satisfaction in product search (to appear in WSDM 2020, in February). AI is a really big deal in China these days. A lot of discussions on the future of AI and what is the next big move. There was a discussion how China should close the gap with the US. All speakers agreed that there is no point in focusing on increasing the number of publications. On the contrary, there are too many low value of those. The Chinese education system (as well as others, I presume) encourages students to publish rather than focusing on meaningful research. According to Michael Jordan (the Berkley’s professor, not the player) Chinese students are the best (the problem, if exists, is with the professors). All you need is just let them work on meaningful tasks. According to Jordan, there is noting in common between DL and AI which handles intelligent decision making. DL which is basically pattern recognition in historical data  cannot replace a doctor who takes decisions for real patients. According to Jordan, the future of AI is in ontologies which are less sexy but capture context. A lot of interesting talks, by Chris Manning on Deep-NLP, Thorsten Joachims on fairness in ranking, Changxhai Yan on the future of IR, and many more. Thank you Jie Tang for the invitation.

Final note, the China’s ban on Google must be stopped. Without gmail, gmaps, gsearch, I felt like blind, dumb, and deaf.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s