tl;dr Began constructing new NLG (and met with Florian, Shruti for feedback)
This week, I finally got to constructing the new (and hopefully improved) NLG, designed by me and based on Shruti's computational model re: giving movie recommendation information. It's of course nowhere near where I'd like it to be, but testing what I currently have with Florian has helped me uncover some "loopholes" and determine new rules to be put in place; and going over the system with Shruti has helped me better understand how to generate accurate and effective responses.
All ten of us InMind team members went down to the Waterfront* for a friendly and fun group dinner, since Yoo Jin and Shruti (the first to leave) are finishing up their internships on Friday.* We had some time to wander the plaza before dinner, and I may or may not have bought myself two books. Also: Yoo Jin, Shruti, Jiachun, and I went back to the Waterfront the following night for a special pre-screening of Crazy Rich Asians.
And finally: since interns' meetings are over for the summer, we filled the time slot with an ArticuLab Field Day--soccer, plus Ultimate Frisbee, plus tag and other silly things at Schenley Plaza!