/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Oriol Vinyals

@oriolvinyalsml
28 posts
2026-02-27
Banana^2 [image]
2026-02-27 View on X
The Keyword

Google rolls out Nano Banana 2, aka Gemini 3.1 Flash Image, with faster image generation, advanced world knowledge, and precision text rendering and translation

Our latest image generation model offers advanced world knowledge, production-ready specs, subject consistency and more, all at Flash speed.

2026-02-26
Banana^2 [image]
2026-02-26 View on X
The Keyword

Google rolls out Nano Banana 2, aka Gemini 3.1 Flash Image, with faster image generation, advanced world knowledge, and precision text rendering and translation

Our latest image generation model offers advanced world knowledge, production-ready specs, subject consistency and more, all at Flash speed.

2026-02-21
Gemini 3.1 Pro has landed! Amazing performance / capabilities across the board. Beyond SOTA, the best are all the things that evals can't measure. E.g. SVG has gotten so much better (see 🧵) https://blog.google/... [image]
2026-02-21 View on X
9to5Google

Google rolls out Gemini 3.1 Pro, which it says is “a step forward in core reasoning”, for all users in the Gemini app; the .1 increment is a first for Google

2026-02-20
Gemini 3.1 Pro has landed! Amazing performance / capabilities across the board. Beyond SOTA, the best are all the things that evals can't measure. E.g. SVG has gotten so much better (see 🧵) https://blog.google/... [image]
2026-02-20 View on X
9to5Google

Google rolls out Gemini 3.1 Pro, which it says is “a step forward in core reasoning”, for all users in the Gemini app; the .1 increment is a first for Google

In November, Google introduced Gemini 3 Pro in preview, with Gemini 3 Flash following a month later.

2026-02-19
Gemini 3.1 Pro has landed! Amazing performance / capabilities across the board. Beyond SOTA, the best are all the things that evals can't measure. E.g. SVG has gotten so much better (see 🧵) https://blog.google/... [image]
2026-02-19 View on X
9to5Google

Google rolls out Gemini 3.1 Pro, which it says is “a step forward in core reasoning”, for all users in the Gemini app; the .1 increment is a first for Google

In November, Google introduced Gemini 3 Pro in preview, with Gemini 3 Flash following a month later.

2026-02-13
An updated & faster Gemini 3 Deep Think is taking off! 🚀 Our smartest mode to date!™️ PhD-level reasoning to the most rigorous STEM challenges (models' gotta think harder). Gold medal-level results on Physics & Chemistry Olympiads. 🧪💻 Full details: https://bit.ly/4kzBLqq [image]
2026-02-13 View on X
The Keyword

Google updates Gemini 3 Deep Think to better solve modern science, research, and engineering challenges and expands it via the Gemini API to some researchers

Our most specialized reasoning mode is now updated to solve modern science, research and engineering challenges.

Fun to see Deep Think's real-world impact. Check out how it's helping researchers catch errors in high-level mathematics research papers. As “just” a math undergrad, I couldn't even dream to do any of this myself! [video]
2026-02-13 View on X
The Keyword

Google updates Gemini 3 Deep Think to better solve modern science, research, and engineering challenges and expands it via the Gemini API to some researchers

Our most specialized reasoning mode is now updated to solve modern science, research and engineering challenges.

2026-02-12
Fun to see Deep Think's real-world impact. Check out how it's helping researchers catch errors in high-level mathematics research papers. As “just” a math undergrad, I couldn't even dream to do any of this myself! [video]
2026-02-12 View on X
The Keyword

Google updates Gemini 3 Deep Think to better solve modern science, research, and engineering challenges and expands it via the Gemini API to some researchers

Our most specialized reasoning mode is now updated to solve modern science, research and engineering challenges.

An updated & faster Gemini 3 Deep Think is taking off! 🚀 Our smartest mode to date!™️ PhD-level reasoning to the most rigorous STEM challenges (models' gotta think harder). Gold medal-level results on Physics & Chemistry Olympiads. 🧪💻 Full details: https://bit.ly/4kzBLqq [image]
2026-02-12 View on X
The Keyword

Google updates Gemini 3 Deep Think to better solve modern science, research, and engineering challenges and expands it via the Gemini API to some researchers

Our most specialized reasoning mode is now updated to solve modern science, research and engineering challenges.

2025-11-19
The secret behind Gemini 3? Simple: Improving pre-training & post-training 🤯 Pre-training: Contra the popular belief that scaling is over—which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is [image]
2025-11-19 View on X
matt shumer

Gemini 3 hands-on: a fundamental improvement on daily use, extremely fast, Antigravity IDE is a powerful launch product, and its personality is terse and direct

Gemini 3 is a fundamental improvement on daily use, not just on benchmarks.  It feels more consistent and less “spiky” than previous models.

The secret behind Gemini 3? Simple: Improving pre-training & post-training 🤯 Pre-training: Contra the popular belief that scaling is over—which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is [image]
2025-11-19 View on X
The Information

Gemini co-lead Oriol Vinyals says Gemini 3's gains come from better pre-training and post-training, contradicting the idea that pre-training gains are falling

which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is [image] Andrej Karpathy / @karpathy : I p...

2025-08-26
We've just upgraded Gemini 2.5 Flash image generation & editing! 🍌🍌🍌 Besides topping leaderboards, it topped my model usage this month. It keeps subjects consistent, you can make precise edits & combine creative elements. Have fun with it @GeminiApp @GoogleAIStudio [image]
2025-08-26 View on X
TechCrunch

Google says it is behind the viral “nano-banana” image model and launches it as Gemini 2.5 Flash Image with finer edit controls in the Gemini app, API, and more

Google is upgrading its Gemini chatbot with a new AI image model that gives users finer control over editing photos …

2025-05-22
Today we introduced Gemini Diffusion⚡️ (& DeepThink, Veo3, Imagen4, 2.5 updates...). It's been a dream of mine to remove the need for “left to right” text generation. It's so fast, that we had to *slow down* the video during the presentation. https://deepmind.google/... [video]
2025-05-22 View on X
The Verge

Google is tapping its users' data to give its AI models an advantage over OpenAI and Anthropic, starting with its opt-in “Gemini with personalization” feature

Google is slowly giving Gemini more and more access to user data to ‘personalize’ your responses.

Today we introduced Gemini Diffusion⚡️ (& DeepThink, Veo3, Imagen4, 2.5 updates...). It's been a dream of mine to remove the need for “left to right” text generation. It's so fast, that we had to *slow down* the video during the presentation. https://deepmind.google/... [video]
2025-05-22 View on X
Fortune

Google DeepMind says Gemini Diffusion, an experimental text diffusion model demoed at Google I/O and available by waitlist, generates 1,000-2,000 tokens/second

Our state-of-the-art, experimental text diffusion model Jose Antonio Lanz / Decrypt : Google Doubles Down on AI: Veo 3, Imagen 4 and Gemini Diffusion Push Creative Boundaries Matth...

2025-05-06
Ahead of I/O, we're releasing an updated Gemini 2.5 Pro! It's now #1 on WebDevArena leaderboard, breaking the 1400 ELO barrier! 🥇 Our most advanced coding model yet, with stronger performance on code transformation & editing. Excited to build drastic agents on top of this! [image]
2025-05-06 View on X
Google Developers Blog

Google unveils Gemini 2.5 Pro Preview I/O Edition with “significantly” improved capabilities for coding and building interactive web apps, before I/O on May 20

We've seen developers doing amazing things with Gemini 2.5 Pro, so we decided to release an updated version a couple …

2025-03-06
Gemini 🤝 Search. Check out AI Mode, our new Labs experiment with the potential to make Google Search drastically more helpful for your hardest questions. AI Mode offers more advanced reasoning, thinking and multimodal capabilities, bringing the best of Gemini 2.0 to Search.
2025-03-06 View on X
TechCrunch

Google rolls out an experimental AI Mode in Google Search, designed to let users ask complex, multi-part questions, to Google One AI Premium subscribers

and you should be, too YouTube : Google AI Mode - The Future Of Google Search? Chris McKay / Maginative : Google Introduces “AI Mode” in Search, Expanding AI Overviews with Gemini ...

2024-12-11
Gemini 2.0 Flash ⚡️ has arrived! 2.0 Flash > 1.5 Pro (again!) 📈 Interacts with a browser 🤖 Native image generation 🖼️ and much more! Try it out https://aistudio.google.com/ ... As a preview of what is possible, wishing you all a Drastic Holiday powered by 2.0! [video]
2024-12-11 View on X
Bloomberg

Google releases its new flagship model Gemini 2.0 and plans to test it in search and AI overviews, saying 2.0 makes “it possible to build agents that can think”

Company expects AI assistants will follow its users around the web  —  Google debuted a new version …

2024-08-03
It was always great fun to discuss ideas and results with @NoamShazeer and @dan_defr Welcome back & let's go! ♊️♊️🚀🚀 https://techcrunch.com/...
2024-08-03 View on X
TechCrunch

Character.AI co-founders Noam Shazeer and Daniel De Freitas, and some employees, are joining Google; Google has signed a non-exclusive agreement to use its tech

In a big move, Character.AI co-founder and CEO CEO Noam Shazeer is returning to Google after leaving the company in October 2021 to found the a16z-backed startup.

2024-05-15
Live from #GoogleIO, we're announcing significant updates to our Gemini family of models. More multimodal, real time, faster, better, longer context... We can't wait to see the amazing things people will do with this tech, as we continue making it more helpful and useful for [image]
2024-05-15 View on X
TechCrunch

Google announces a private preview of a new Gemini 1.5 Pro version that can take in up to 2M tokens, twice its predecessor and rivals like Anthropic's Claude 3

Gemini, Google's family of generative AI models, can now analyze longer documents, codebases, videos and audio recordings than before.

Live from #GoogleIO, we're announcing significant updates to our Gemini family of models. More multimodal, real time, faster, better, longer context... We can't wait to see the amazing things people will do with this tech, as we continue making it more helpful and useful for [image]
2024-05-15 View on X
Engadget

Google announces Gemini 1.5 Flash, which is more lightweight and cheaper than Gemini Pro, but has the same multimodal capabilities and 1M-token context window