Fitzcarraldo — or How to Hack Academia to Build Stuff
AMAsupported by Facebook
What does a movie about a monomaniacal quest to build an opera house in the Amazon Basin has to do with science? Programming language research is at a crossroads. While formal language techniques are advancing steadily, the transfer of ideas to practice is much slower. Applied research is based on experimentation which is becoming prohibitively onerous. To evaluate any novel idea requires dealing with the complex language ecosystems of the day. Even the simplest innovation may take years of work to try out. Without a realistic evaluation, papers are rejected, and underlying hypotheses remain untested. Academic researchers are in a particularly precarious position. On the one hand, they must compete with industrial R&D teams with massively more resources, and on the other hand, they are beholden to an incentive system that rewards copious publication rates at select conferences. How can we build stuff that matters in such an environment? Are we doomed to failure? In this talk I argue that trying to build stuff is the necessary starting point, failure is expected and should be embraced, as, without it, we are not working on anything that matters. Based on my experience it is possible to navigate the academic system and come up with ideas that occasionally matter.
This talk is given by the recipient of the Dahl Nygaard Senior prize.
The AMA following this talk will be moderated by Sophia Drossopoulou.
Sun 15 NovDisplayed time zone: Central Time (US & Canada) change
07:20 - 08:20
|Fitzcarraldo — or How to Hack Academia to Build StuffAMAsupported by Facebook|
Jan Vitek Northeastern University / Czech Technical UniversityLink to publication Media Attached