So am I understanding this correct: You are here for a $1600 conference. And it’s not for work. This is your free time?
I recently travelled to San Francisco to attend the Google Next conference. Clearly neither a scientific conference nor related to geophysics. And also clearly, the immigration officer was not particularly on board with my choice of leisure activity.
In total, I spent a week in SF, from which, one day was dedicated to the Google Next conference and two days were allocated for Kaggle days. The rest was discovering the centre of modern tech culture. Was it worth it, travelling there on my own dime?
Well, yes. Yes, it absolutely was. Let’s start at:
Google Next ’19
Deemed the cloud conference, the keynote in the early hours of the conference included gems like Anthos, a tool to deploy jobs and apps to “any public cloud”. The speaker was clearly briefed not to say AWS. But it seems like a great technology to avoid vendor lock-in. But mainly it was impressive, from which angles Google approaches the cloud business. But see for yourself:
Personally, my favourite was the talk on Tensorflow 2.0 by Paige Bailey and Lak Lakshmanan. Clearly, because I’ve been using TF for most of my work and personally, I’ve been following Paige on Twitter for a long time. She’s a geoscientist by background, that progressed to work at Microsoft and went on to work to bring TensorFlow into the future. TensorFlow 2.0 includes Keras natively in the TensorFlow architecture and got rid of some huge nuisances that other deep learning libraries did not have. (Yes TensorFlow does Eager execution now.)
I was a bit lucky after. Not sure if I was intruding, but Paige Bailey met with the TensorFlow testers after and I got some insights into what they’re up to. Extremely accommodating folks and I was glad to spend that time.
So straight from the horse’s mouth:
There’s a tensorflow hackathon going on, to bring projects to TF 2.0
You may recognize the face in the video introducing you to the challenge.
There is also an event focusing on documentation coming up. Generally, the crowd seemed to welcome every contribution, especially to improve the documentation.
That is also when I met Andy Ferlitsch. A Googler that was hanging out at the Community Hub. We had a chat about deep learning and he even showed interest in my Ph.D. work! I got some fantastic feedback on the latest paper I’m writing on deep learning CNNs.
All the talks are available on Youtube. But the experience and mingling with all these interesting people was honestly priceless.
In the evening I got to hang out with the first few Kaggle Days participants. A fantastic event in and of itself. Please subscribe to receive that update right when I hit publish!
Latest posts by Jesper Dramsch (see all)
- Interactive Viz² and Differentiable Machine Learning – Friday Faves - 2019-11-01
- Dear EAGE we have to talk! - 2019-10-30
- I like oxygen – Explaining #teamtrees - 2019-10-26