Last Friday was my last day at @GoogleDeepMind.
4 years ago, I joined Google to teach LLMs to call APIs. I left having worked across some hardest parts of Gemini's evolution into an agentic model, on the very team that created BERT, a model I was training on top of before I joined Google.
Proud of what I shipped here, with great colleagues along the way:
- captained multiple flagship Gemini releases
- unified tool-use protocol
- pushed the frontier of long-horizon RL across Computer Use and complex tool use
- improved Search-enabled response quality at scale
- built data flywheel to optimize data distribution across training stages
Grateful to every person who made the last 4 years what they were, including those who dragged me away from the screen once in a while.
A few days unaffiliated. More soon.
Joint Statement: Apple and Google have entered into a multi-year collaboration under which the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology. These models will help power future Apple Intelligence features, including a more personalized Siri coming this year.
After careful evaluation, Apple determined that Google's Al technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users. Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards.
@aclmeeting Do we have a "Language Generation" track this year? I didn't see that listed on the conference website https://t.co/dVL6vjJqkC, but found it on Softconf submission page.