3 ms·
Be serious. Everything ML/AI at Google post-dates 2010, including 4 generations of TPUs. You have made the common mistake of believing that Google's products ar
by jeffbee 5y ago
Be serious. Everything ML/AI at Google post-dates 2010, including 4 generations of TPUs. You have made the common mistake of believing that Google's products are the things in the hamburger menu, but Google's actual product is dirt-cheap computing.
- endisneigh 5y agoEven by this definition Google isn't the cheapest nor the most available in that time period. Even you're claiming Google's actual product is dirt cheap computing then they're failing by price, market capture percentage and breadth of offerings.
- dekhn 5y agoNo. Early google had several machine learning systems including Phil, RePhil, SERI, SmartAss (which made google many billions before 2010), and Sibyl all of which predate 2010. Rephil: https://www.quora.com/How-does-Googles-topic-model-Rephil-work https://www.quora.com/How-does-Googles-topic-model-Rephil-wo... and 26.5.4 Case study: Google’s Rephil 928 from Kevin Murphy's ML book. Rephil was critical for google when they needed to cluster topics and it was used for a wide range of other things. Smartass: http://static.googleusercontent.com/media/research.google.com/en/us/pubs/archive/41159.pdf http://static.googleusercontent.com/media/research.google.co... Smartass is really weird if you come from the "batch training of deep networks" world, because it's online, and visits the dataset once, in temporal order. But. it's as ML as it gets; it was just turning joined clicks (between search queries and ad impressions) into weights for pCTR. Sibyl was Google's secret sauce before tensorflow started in 2007, it implemented parallel boosting using mapreduce as a worker engine to do linear and logistic regression and was used to train the models for youtube watch next and play store (both of which produced enormous revenue at a time when people had no idea that youtube and android were gold mines) but predated those uses by at least 5 years. It was one of the oddest systems I've ever worked on and you can see some of the ideas it innovated in TFX today. There's also SETI which IIRC even predated Sibyl, and you still can see vestiges of its functionality when doing archeological work on the google codebase. I worked on systems related to smartass, worked on sibyl, and later tensorflow/TPU hardware/software. I met a number of long-time engineers and asked them deep questions about ML use at Google predating tensorflow and gained a deep understanding (and even then, I've left out many details about systems I'm not aware of). Before tensorflow there wasn't much use of GPUs or TPUs, for several reasons but once Vincent Vanhouke stuffed a few GPUs in his workstation and showed fast training to Jeff Dean, that changed quickly.
- acdha 5y agoML/AI is definitely big but they're not really changing the world with it. Adding it to Photos has been nice but hardly critical (search is handy but “enhancements” just meant I had figure out how to disable it) and adding it to Search has been at best neutral. GCP has some good stuff but it's not especially compelling compared to the alternatives which have more features. I wouldn't be as negative as the original poster but they do seem poorly led and that's hurting them. Even GCP gets hit with C-level concern that “Google kills products” and their sales team has been under-motivated in my experience.