Hey HN! Here's the rundown:<p>- Exabyte crowdsources compute from Exabyte users.<p>- Exabyte uses crowdsourced compute to train deep learning models and advance research.<p>- I like to think of this crowdsourced compute model as BitTorrent for researchers.<p>- I'm a researcher frustrated by existing compute options for training deep learning models and want to create something new.<p>- This is totally a work in progress. All feedback is welcome :-)