50 karma · joined July 28, 2019
Google does have an experimental API, but have not found an associated blog post or paper with it: https://cloud.google.com/ai-workshop/experiments/generating-...
So overall 10-20 hrs a week, with the third outlet consuming the maximum time
I love some of the regular shows there like GoodTimes, deep learning research review hosted by Andrej Karpathy etc. When I step into smaller rooms, the signal to noise ratio decreases (on topics I care about).
I felt clubhouse to be more conversational than podcasts (which are often a Q&A format). It also allow new people to participate, ask questions.
Exploration of topics and rooms is pretty hard currently. How do you discover interesting people and conversations? For me it has mostly been twitter, which is non-optimal. My second big pain point has been the FOMO nature of it. I would ideally like to consume content on my time and schedule not the other way around.
For some domains, with privacy concerns or rarity of objects, getting labelled data for deep learning is challenging.
There is decent research on sim2real i.e transferring models trained on synthetic data to real world applications https://arxiv.org/pdf/1703.06907.pdf