I just reread Breiman’s “Two Cultures” after many years. It remains an absolute gem of a paper, easy to read with minimal equations. And it’s probably even more relevant today than when it was written in 2001. If you haven’t read it, you should: https://t.co/W3If8b0Mlm
— David Smith (@revodavid) August 9, 2019
At one end of the spectrum, we have AI companies that do incredible work on billions of $.
— Chip Huyen (@chipro) August 9, 2019
At another end, we have teams like @spacy_io (2 pax) & @huggingface (3 pax) that also do incredible work on single-digit GPUs.
The ability to advance AI field is resource-independent.
I was really prepared to hate this article on HN (hyperbolic examination of the political angle that's so common these days), but this description of the tone of the site is 100% spot-on. https://t.co/teyRr1BKMO pic.twitter.com/QLgji91EGH
— Vicki Boykis (@vboykis) August 8, 2019
Great, detailed history on the role of surveillance in trying to suppress movements for social change, and how privacy makes advocating for civil rights possible (from the Pilgrims to Cesar Chavez to the current day, & more), by @alvarombedoya https://t.co/37XICmSW1n pic.twitter.com/StKzuQmgpd
— Rachel Thomas (@math_rachel) August 7, 2019
Sorry, what I wrote was too simplistic. The authors find that the impact of papers has a low correlation with the number of papers and impact is independent of when you publish (so close to zero auto-correlation). Streaks of great publications are often luck (or a sign of high Q)
— Tim Dettmers (@Tim_Dettmers) August 7, 2019
Client: We'd like to do what big AI labs like OpenAI and DeepMind are doing.
— Chip Huyen (@chipro) August 7, 2019
Me: You mean getting into a $1B debt?
Client:
Me:
Client: So is there any free pretrained model that you suggest that we use?
There's some links here: https://t.co/3syIJtnHsN .
— Jeremy Howard (@jeremyphoward) August 7, 2019
This is why data provenance is so important. 😂 https://t.co/y7Zhc0vnwM
— Hilary Mason (@hmason) August 7, 2019
There's actually some s/w eng research that shows no correlation between years of experience and programmer productivity!
— Jeremy Howard (@jeremyphoward) August 7, 2019
After working for a while I realize how hopelessly outdated the concept "years of experience" is. It's so easy to spend years on a job and learn absolutely nothing.
— Chip Huyen (@chipro) August 7, 2019
how cookies work pic.twitter.com/OJyLAj0Xp5
— 🔎Julia Evans🔍 (@b0rk) August 6, 2019
We've analyzed the results of the Dask User Survey: https://t.co/GOaSHF8COE
— Dask (@dask_dev) August 6, 2019
Or run your own analysis on @mybinderteam: https://t.co/Xn1rm3KgsD
Thanks to everyone who participated!
