A good team isn't one that doesn't have disagreements, it's one where disagreements are possible and productive.
— Smerity (@Smerity) October 3, 2018
A good team isn't one that doesn't have disagreements, it's one where disagreements are possible and productive.
— Smerity (@Smerity) October 3, 2018
“[AutoML] sells a lot of compute hours so it’s good for the cloud vendors” - @jeremyphoward
— Peter Skomoroch (@peteskomoroch) October 2, 2018
Adapting best software development practices for training and using deep networks by @karpathy:
— Alexandr Kalinin (@alxndrkalinin) October 2, 2018
- careful development of the test set to track edge cases
- train and deploy faster, test automatically
- version control datasets and models#pytorch #ptdc2018 pic.twitter.com/aP6aHfW95M
Glorious post from @talyarkoni on the "appeal to incentives" fallacy in scientific practice, which I admit I've been guilty of deploying myself. This post will make you think and reconsider https://t.co/ARXQaj4gt1 pic.twitter.com/zO1gFOerRX
— Chris Chambers (@chrisdc77) October 2, 2018
@fastdotai fellow Sanyam asked me a few questions about my background, doing research, & writing articles. I hope the answers may be helpful to some people. https://t.co/uYdE40cjLt
— Sebastian Ruder (@seb_ruder) October 2, 2018
Or inversely - work on privacy preserving tech so you can train on big data to which you might not have plaintext access
— Trask (@iamtrask) October 1, 2018
100x this. Be respectful, be careful, be correct. https://t.co/PlOCRtUhj1
— Chris Albon (@chrisalbon) September 30, 2018
We'll make better choices when we keep this in mind https://t.co/3qhNjOKder
— dj patil (@dpatil) September 30, 2018
“We need a new generation of AI researchers who are well versed in and appreciate classical AI, machine learning, and computer science more broadly while also being informed about AI history.” @AdnanDarwiche, on why deep learning is not enough https://t.co/RFAR3JoXOl
— Gary Marcus (@GaryMarcus) September 30, 2018
I implore young ML-ers to read abt the breaches of integrity that sunk Brian Wansink: "misreporting of research data, problematic statistical techniques, failure to properly document and preserve research results, and inappropriate authorship" https://t.co/EZplGyLq9h
— Zachary Lipton (@zacharylipton) September 30, 2018
For many companies, transitioning to mobile required using a pre-existing API. Unobtrusive to the rest of infra. Machine learning promises to lay siege across all the layers of your infra stack, good idea or bad. Mobile and ML love novel UIs to make new capabilities accessible. https://t.co/WUl9V5GE8l
— Smerity (@Smerity) September 29, 2018
Empathy is everything. https://t.co/ZpiBXDarjG
— 👩💻 DynamicWebPaige @ #APICityConf 🌇 (@DynamicWebPaige) September 29, 2018