Starship and Super Heavy are ready at the launch pad in Starbase, Texas. Targeting Saturday, November 18 for Starship’s second integrated flight test → spacex.com/launches

🚨 [New Paper] If you're involved in AI safety or jailbreaking, you don't want to miss this: Techniques from human communication now effectively breach aligned LLMs (Llama-2 Chat, GPT-3.5, GPT-4) with over 92% attack success rate. 👇🧵(1/7 - page link: chats-lab.github.io/persuasive_jai…)

Formula 1 x Apple Vision Pro This will change the sport forever Links below 👇

Our vision for Act II of Arc

The number one reason why good security is hard is that the feedback loop on decisions is long and the signal is low fidelity. It's not clear how many incidents were prevented or mitigated from which foundational decisions years prior. This wrecks the incentives to be proactive.

Self-replicating prompt injection w/ Ben Nassi wired.com/story/here-com…

Starship completed its rehearsal for launch, loading more than 10 million pounds of propellant on Starship and Super Heavy and taking the flight-like countdown to T-10 seconds

announcing... starlinkmap dot org real-time map of every starlink satellite. tracks upcoming launches, other constellations, orbital updates, etc. finally launching this after a while! more details below.

Congrats Google DeepMind on the new Gemma-2 27B & 9B release! Gemma-2 was tested in the Arena under the codename "*late-june-chatbots" and now out of stealth. Its early result matches the best open models (Llama-3-70B, Nemotron-340B) with only 27B parameters! Impressively,

Advanced mathematical reasoning is a critical capability for modern AI. Today we announce a major milestone in a longstanding grand challenge: our hybrid AI system attained the equivalent of a silver medal at this year’s International Math Olympiad!

Exciting News from Chatbot Arena! Google DeepMind's new Gemini 1.5 Pro (Experimental 0801) has been tested in Arena for the past week, gathering over 12K community votes. For the first time, Google Gemini has claimed the #1 spot, surpassing GPT-4o/Claude-3.5 with an impressive

Yeah, Gemini-exp-1114 is pretty good :)

Gemini-exp-1206, our latest Gemini iteration, (with the full 2M token context and much more) is available right now for free in Google AI Studio and the Gemini API. I hope you have enjoyed year 1 of the Gemini era as much as I have. We are just getting started : )

What a way to celebrate one year of incredible Gemini progress -- #1🥇across the board on overall ranking, as well as on hard prompts, coding, math, instruction following, and more, including with style control on. Thanks to the hard work of everyone in the Gemini team and

Blown away by our new image and video models. Glad to see others are as well! The mix of creativity and realism is 🤯

As we make progress towards AGI, developing AI needs to be both innovative and safe. ⚖️ To help ensure this, we’ve made updates to our Frontier Safety Framework - our set of protocols to help us stay ahead of possible severe risks. Find out more → goo.gle/42IuIVf

Introducing Gemini 2.5 Pro, the world's most powerful model, with unified reasoning capabilities + all the things you love about Gemini (long context, tools, etc) Available as experimental and for free right now in Google AI Studio + API, with pricing coming very soon!

Deep Research in the Gemini App is now powered by Gemini 2.5 Pro, and our early tests show users prefer this 2:1 vs “other products” ;) gemini.google.com

Per our Frontier Safety Framework, we continue to test our models for critical capabilities. Here’s the updated model card for Gemini 2.5Pro with frontier safety evaluations + explanation of how our safety buffer / alert thresholds approach applies to 2.0, 2.5, and what’s coming.

cooking up something tasty for tomorrow...