My 5978 Likes

✨ Introducing Gemma 3n, available in early preview today. The model uses a cutting-edge architecture optimized for mobile on-device usage. It brings multimodality, super fast inference, and more. https://t.co/EgoQlyhzEo

Introducing Flow: a new type of AI filmmaking tool that combines the best of Veo, Imagen and Gemini — built with and for creatives. Flow helps you maintain character and visual consistency from one clip to the next. See how emerging filmmakers are using it 🎥 https://t.co/H0cBv6IGs1

Get into the zone with Flow. 🎬 It combines the best of our most advanced models Veo, Imagen and Gemini into 1️⃣ master filmmaking tool - helping you weave cinematic clips, dynamic scenes, and compelling narratives into stories with consistent results. https://t.co/E2a7NJNuP9

Your toughest questions. Your deepest research. Your personalized shopping needs. Your just-save-me-time tasks. All you have to do is ask. Welcome to the next chapter of Google Search. #GoogleIO https://t.co/gZoejJoZM5

We’ve developed Gemini Diffusion: our state-of-the-art text diffusion model. Instead of predicting text directly, it learns to generate outputs by refining noise, step-by-step. This helps it excel at coding and math, where it can iterate over solutions quickly. #GoogleIO https://t.co/EmDAqNeXVv

Last year we unveiled Project Astra on the #GoogleIO stage. See how it’s evolved since then — and what might be possible in the future. https://t.co/ObMi7gFrrl

Last year, we introduced Project Astra: a research prototype exploring capabilities for a universal AI assistant. 🤝 We’ve been making it even better with improved voice output, memory and computer control - so it can be more personalized and proactive. Take a look ↓ #GoogleIO https://t.co/MziQ0KuN0G

hi. my name is vanes. i'm 33 and completely lost. but here's my best bet at figuring it out. https://t.co/ZxsX70ulFl

1. Build II-Agent, possibly the best agent in the world 2. Open source it ... 4. Everyone profits Our mission at II is to increase humanity's intelligence II-Agent is the next part of the stack we are building to enable this, topping the agent benchmarks 🧵 https://t.co/6ri8xDZNlz

GPS isn’t broken. It’s just vulnerable. Today, we’re proud to share a glimpse into what we’re building at Dirac Labs. A quantum sensor that listens to the Earth itself — no satellites required. Spoof-proof. Jam-resistant. Always on. 🌐 https://t.co/bPjGRj0lDX https://t.co/5oO0WbtmKZ

🇮🇩 See how AI is transforming Indonesia – LIVE! #GoogleCloudSummit Jakarta is almost here. And we'll be livestreaming the keynote on May 22 right here on X. Like ❤️ this post for a reminder to join in when we're live. https://t.co/57PLc8lD6T

In India 🇮🇳, students in colleges prepare to enter the real world by doing fake things: - meaningless resume padding projects - endless interview prep - placement politics We can do better. Introducing @quest_sh: A set of hard but tractable *real world* problems you can solve… https://t.co/zwMAqJU5wW

An Imperial College eng professor gave four LLMs a problem set that graduate students had two months to solve. He had TAs grade the results blind alongside real submissions. Meta AI and Claude failed. ChatGPT ranked 27 of 36 students...while Gemini 2.5 Pro ranked 4 of 36 🤯 https://t.co/VcLt4d0MnW

The best part about making this template is that I can finally tell my parents I used my Philosophy degree for something. https://t.co/ZAzygqGKVC https://t.co/w7upZaxv4f

What will humans be doing in 10 years? 1. In a decade most people will not be driving. What will we do instead? 2. In a decade most people will have robots in their businesses and probably their homes. 3. In a decade we won't be looking at 2D monitors much anymore. 4. In a…

Science shows that you need to be extreme in order to get what you want. Video by Emily Mc Donalds https://t.co/xQod3Gv1ff

whenever I see someone level up really fast, it’s usually one of three things: 1. confidence catches up to competency and they stop second-guessing 2. accepting who they actually are 3. environment switch to faster moving water

It's remarkable that Douglas Adams was the only scifi author who came within a mile of predicting the behavior of actual AI https://t.co/161xxpAybI

Thought I might share a life story from a few months ago to brighten your day. Back in March, I spotted a raccoon on the side of the interstate that had very clearly been run over. I pulled over thinking I’d have to put it out of its misery. We ended up becoming good friends. https://t.co/rdWahB5RJu

Introducing: Infinite Chat API 💥 Extend the context length of ANY model ...while saving 90% of your tokens and cost AND improving performance! One single line to switch. Available to use now https://t.co/hDfAuu2skD

"I've figured out how to compress everything I've learned into one biomarker." - @bryan_johnson "It's your resting heart rate before bed. If you show me your resting heart rate, I'll see your soul. It's an accounting of your life decisions." "A lot of our worst behaviors are… https://t.co/ssdy2M28lk

Introducing AlphaEvolve: a Gemini-powered coding agent for algorithm discovery. It’s able to: 🔘 Design faster matrix multiplication algorithms 🔘 Find new solutions to open math problems 🔘 Make data centers, chip design and AI training more efficient across @Google. 🧵 https://t.co/y5QQNpKm0L

🎉 Introducing CodeRabbit for VSCode, Cursor, and Windsurf 🎉 Vibe check your code where you write it today! (Link below!) https://t.co/H2I1iW8Eng

Constructor Theory of Time: https://t.co/FgsECTHvTt

Was just getting warmed up https://t.co/dN6rSXiLVS https://t.co/JbJ8TxyaAz

I’ve gotten am embarrassing amount of leverage out of “I’m afraid. I guess I’ll just do it while afraid” and “this is going to take a long time, but the time is going to pass anyway”

I spoke at a high school yesterday and someone asked what I'd tell my 18 year old self if I could. It would be not to be tricked by prestige. It is at best a trailing indicator, and often simply wrong. Curiosity is a better guide.

In September, 2024, physicians working with AI did better at the Healthbench doctor benchmark than either AI or physicians alone. With the release of o3 and GPT-4.1, AI answers are no longer improved on by physicians. Also error rates appear to be dropping for newer AI models. https://t.co/HI2GBXiLS8

What if humanity forgot how to make CPUs? Imagine Zero Tape-out Day (Z-Day), the moment where no further silicon designs ever get manufactured. Advanced core designs fare out very badly. Assuming we keep our existing supply, here’s how it would play out: https://t.co/Fy1QksSPIF

Evaluations are essential to understanding how models perform in health settings. HealthBench is a new evaluation benchmark, developed with input from 250+ physicians from around the world, now available in our GitHub repository. https://t.co/s7tUTUu5d3