Current Computer Science Masters student at the University of Warwick.
Joined May 2011
- Tweets15
- Following120
- Followers78
- Likes46
Sean Hooker retweeted
A delightful essay by @GoogleAI's @sarahookr on the notion that often the success of research idea depends on the idea and how well it suits the available HW and SW, rather than just the merits of the idea. It's an enjoyable and read, Sara!
Really great history + position paper by @sarahookr:
The Hardware Lottery (arxiv.org/abs/2009.06489)
"This essay introduces the term HW lottery to describe when a research idea wins b/c it suits the available SW/HW, not because the idea is superior to alt. research directions."
Sean Hooker retweeted
Welcome back, today we are thrilled to have @NailaMurray join the podcast.
What does it means for a picture to be "interesting" to a human? Also, what do optical illusions teach us about the brain.
@SeanHooker gets to grips with data privacy.
Enjoy!
bit.ly/3kNryan
Sean Hooker retweeted
Welcome back, today @julius_adebayo joins the podcast!
Julius brings a fantastic question to the table "Can a neuroscientist understand a processor?"
We also revisit a 1998 prediction that computer hardware would match the human brain by 2020.
Enjoy!
bit.ly/3a29K6D
Sean Hooker retweeted
We are back with Season 2!
An exciting premier featuring
@huangcza
We talk about our highlights from season 1, judging the Eurovision AI song contest and the use of ML to predict child welfare on crisis hotlines.
underratedml.com
Sean Hooker retweeted
Underrated ML is back!
@Smerity joins the podcast and we talk about strongly type cast recurrent neural networks, borrowing ideas from physics and functional programming.
We also discuss Turing's 1952 paper on how patterns emerge in nature. How did the leopard get its spots?
Sean Hooker retweeted
underrated ml is back!
@seb_ruder joins the podcast to discuss the challenges of linguistic diversity and inclusion.
There are over 7,000 languages in the world, why does ML research overfit to a small subset?
underratedml.com/
Sean Hooker retweeted
Docking confirmed! @AstroBehnken and @Astro_Doug officially docked to the @Space_Station at 10:16am ET:
Sean Hooker retweeted
So enjoyed this chat with @le_james94. He puts so much effort into each episode — I am amazed at the effort he took to prepare such detailed questions!
Plenty of fun q’s about my unusual upbringing and academic background + ml research on interpretability and model compression.
(1/19)
The 32nd episode of #Datacast is my conversation with @sarahookr, deep learning researcher at @GoogleAI and founder of @deltanalytics. We broadly discussed her economics background, data for good, and her AI research.
datacast.simplecast.com/epis…
Sean Hooker retweeted
i’m worried i babbled pretty incoherently throughout the podcast, and i ask for your kind understanding: it was just so much fun talking with @sarahookr & @SeanHooker and i kinda let go of any restraint i usually have
Underrated ML is back!
@kchonyc joins the podcast and talks about some underrated work on energy functions -- why it may be sensible to return a set of plausible and implausible predictions.
We also discuss remote organizing of ICLR, shortcut learning and recent NLP progress.
Sean Hooker retweeted
Underrated ML is back!
@kchonyc joins the podcast and talks about some underrated work on energy functions -- why it may be sensible to return a set of plausible and implausible predictions.
We also discuss remote organizing of ICLR, shortcut learning and recent NLP progress.
Sean Hooker retweeted
Thanks @reworkdl for profiling our little podcast along with some other fantastic listens. :)
Top AI & Data Science Podcasts: bit.ly/2YKTEug
Listen now: underratedml.com/
Sean Hooker retweeted
Do you know what it's like to be asked on a show that you are a fan of? Well now I do! 🥳
My podcast debut! So much fun and lots of laughters, plus some serious discussion of science! Thanks for having me and asking me questions I can't stop thinking about 🤔
We have @savvyRL on the podcast later today.
We prompted gpt-2 to generate some questions for Rosanne. Here is one gem it came up with:
"What is your favorite idea from the IPCC 5th Assessment Report (AR5) or from one of the blogs you follow?"
What else should we ask? :)
Sean Hooker retweeted
Rosanne Liu joins the podcast, and talks about the beauty of a thorough research treatment of small ideas -- like why have modern deep neural networks stopped using pooling layers?
Does pooling actually deliver on translation invariance?
bit.ly/2xquBkX
Sean Hooker retweeted
Drumroll for our new website...
underratedml.com/
Cozy at home + want to take your mind off of covid-19?
Our first two homespun podcast episodes are live.
