Oh we're doing more fun BTS here? I accidentally invented generative retrieval as a side project because my manager at Google at that time was very into IR/retrieval but IR was kind of a sunset/boring field so I kind of wanted to make it cool again After DSI was born,
GENERATIVE AI
-
Gemini 2.5 Flash Thinking impresses with quality-cost trade-off
By
–
We've started shifting more to Gemini 2.5 Flash Thinking, and 2.5 Pro and it's impressive. Flash Thinking in particular has been ground breaking in the quality-cost trade off.
-
SAS Ready-Made AI Models for Enterprise Innovation
By
–
Looking at our ready-made AI models like: #SASInnovate
-
Gen-4 References Update Enhances Quality and Scene Composition
By
–
We've released an update to Gen-4 References that brings marked improvements to aesthetic quality, scene composition and identity preservation. Alongside this update comes a number of exciting new use cases which we'll be sharing more about in the coming days. Gen-4 References
-
GenAI Adoption: From 98% Use to Real ROI Strategy
By
–
By 2028, 1B+ apps will be built using GenAI.
Yet today, 98% of enterprises use AI – but only 25% see full ROI. The gap? It’s not the tech – it’s the strategy. Hybrid infrastructure, connected data, and automation are key to scaling real value. Want to go deeper? Tune into -
Copilot Adds Edit Response Feature with Saved Pages
By
–
Long-press/hover on a Copilot response (new or old), hit "edit this response" and it'll create a Page. Edit it yourself or ask Copilot to add more info/refine what's there until it's exactly what you want. Saved right above chat history for easy access.
-
PyMuPdf drops figures; keeping images requires switching to VLM
By
–
Just PyMuPdf! So figures are dropped, a good direction for improvement would be to keep images, but then we'd need to switch to a VLM instead of LLM.
-
Gemini 2.5 Pro Achieves State-of-the-Art Video Understanding
By
–
Gemini 2.5 Pro (05-06) is SOTA at most video understanding tasks (by a large margin) . Lots of work by the Gemini multimodal team to make this happen, excited to see developers push this capability in new ways. More details below!
-
Multimodal AI team releases YouTube video understanding feature
By
–
Congrats to @AniBaddepudi @AntoineYang2 @MarioLucic_ @jalayrac + the rest of the MM team on this! And in AI Studio and the API, you can paste a YouTube link and test this very quickly : ) Learn more:
-
Using o3 Prompting for Automated Curriculum Design
By
–
Steal my o3 prompt to create a comprehensive 30-day course to learn and master any topic ——————————————-
ULTIMATE 30-DAY MASTERY ROADMAP ——————————————- ULTIMATE LEARNING ROADMAP ARCHITECT: 30-DAY MASTERY SYSTEM
You are a