Machine learning at Google
summary
Google describes real deployments: search shifts from ~two and a half words of query context toward conversational inputs after 13 years of ML work, cutting rule-writing that took years to three to four months to get a first model. Cloud services show phone AR translation into 103 languages, self-driving stacks with a few million miles of driving data, and infrastructure that can scale to hundreds of thousands of machines per prediction. TensorFlow is the open option; hyperparameter tuning and production tooling remain the hard engineering work.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:47Machine Learning is...
- 1:27Enhanced Search
- 5:12Keys to Successful ML
- 5:52Offerings across the spectrum
related talks
59mMachine Intelligence at Google Scale
27mKeynote - Google Cloud, Machine Learning
10mMachine Learning: The next great leap
42mMachine Learning APIs, from Training [...]
47mBuilding smarter apps with Machine Learning, from magic to reality
13mTaking Machine Learning to the Next Level
32mMachine Learning with Google Cloud Platform
57mBuilding smarter solutions with Machine Learning, from magic to reality