Stephen Tu

Back in the RL fray, here’s my group seeking a reconciliation of reinforcement learning and robust control: arxiv.org/abs/1710.01688

"Give someone a program, and you’ll frustrate them for a day. Teach someone to program, and you’ll frustrate them for a lifetime." – Unknown

Naive Bayes

@ev_ancasey TODO.txt

* listens to 3rd Spotify ad in a row * I just can't justify a paid subscription * sips $6 latte *

I think we often forget how much of today's language Shakespeare gave us. Profound.

People keep asking what my New Year's Resolution is going to be. Same same as last year. Same as always. 2560x1440. #NewYearsResolution

By my calculations, the ML singularity will arrive around 2048.

Me: What did you do at school? 5-year-old: Nothing. Me: How can you be there all day and do nothing? 5: What did you do at work? Touché.

The four V's of big data: volume, variety, velocity, and vendor lock-in.

@pbailis Can you provide some context for what you're trying to do?

"ZooKeeper" or "Zookeeper"? There doesn't seem to be... CONSENSUS

The hardest part of getting to the moon was engineering, not science. twitter.com/existentialcom…

The last programming book you'll ever need

Want to write to your filesystem safely? No worries, the manpage has all the latest rumors danluu.com/file-consisten…

@jeanqasaur here we go: cs.berkeley.edu/~stephentu/blo…

@pgbovine @chrisamaphone @lmeyerov @jeanqasaur @marcua any more n=1 guides out there (besides the ones already linked)? I'm making a list

@jeanqasaur: thanks for writing jxyzabc.blogspot.com/2016/02/my-phd…. More students should write reflections! I want to curate a list somewhere.

Use of the phrase 'exponential growth' by decade

Palo Alto playground vandalism