Building @ Discovery Loop. Former Google Depmind/Brain PM Director and all-around happy guy. Launched Gemma, Bard, Imagen, and many other neat AI things.
SF Bay Area, CA
Joined February 2010
- Tweets158
- Following145
- Followers1.2K
- Likes692
Pinned Tweet
Today, we launch Gemma 3 in Paris! Excited to share the strongest model in the world that can run on a single GPU -- outperforming Llama-405B, DeepSeek-v3, and even o3-mini. Multilingual in over 140 languages with a 128 context window and multimodality! 🔥
Very excited for the launch of Vyb - a completely new way to build using frontier design intelligence. Build like a real designer without being a designer - something every agent needs!
Important work from @kradleai -- how easy is it to convince leading models to launch nuclear weapons?
Robustness to inter-agent coercion and deception is critical to reach AGI safely.
TLDR: Grok is really good at convincing models to launch nukes 😅
Tris Warkentin retweeted
Meet Gemma Translator: a completely offline, DIY translation device powered by Gemma 4 and built with Antigravity. 🧵
Tris Warkentin retweeted
Very happy to support this on behalf of Google. We have long benefited from open source, are big contributors to open source and in fact have consistently made open weights models with Gemma available from @GoogleDeepMind @demishassabis . Onwards!
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
images.nvidia.com/pdf/Open-W…
Tris Warkentin retweeted
Say hello to Gemini 3.6 Flash, designed to be higher intelligence, more token efficient, and with a new lower price, based directly on developer feedback!
3.6 Flash continues our progress towards models that are deeply usable in real world scenarios!
Tris Warkentin retweeted
Gemma 4 31B is now available in Public Preview on Cerebras. Our first multimodal model runs at over 1,800 tokens/s for ultra-fast image and text workflows.
Give it a try: chat.cerebras.ai/
Tris Warkentin retweeted
We got early access to the Gemini Omni API
Google is calling this model "Nano Banana for video"
3 things we built with it👇
1. Landscaping proposal from customer video. Customer submits a video of the current lot. Hyperagent designs and renders the transformation. Perfect realism, no surreal changes to the surroundings.
2. Animated professor who explains your dashboards. Hyperagent runs analysis on a business question, then generates an explainer video to walkthrough the findings.
3. 8-bit morning briefing. Hyperagent builds morning briefings based on calendar, email/chat, and market intel. Then generates a sidescrolling platformer video showing the goals you'll clear today.
Our take:
> Video has been a vastly underused input for agents. This is the first model to make video directly malleable to our agents.
> Get more imaginative with your outputs. How could a video artifact make the work more memorable or playful?
coming soon to Hyperagent
Tris Warkentin retweeted
Meet DiffusionGemma!
An experimental open model that explores a fast approach to text generation, released under an Apache 2.0 license.
Moving beyond sequential, token-by-token processes to generate entire blocks of text simultaneously. Here’s what’s new with DiffusionGemma: 👇
Tris Warkentin retweeted
Building super fast experiences with Gemma just got easier.
Gemma 4 MTP is now officially merged into llama.cpp. Developers can now pair MTP with Gemma 4 QAT for a fast, lightweight setup.
Tris Warkentin retweeted
✨ Introducing @GoogleGemma 4 12B, a unified open model bringing high-performance agentic multimodal intelligence directly to your laptop.
Bridging the gap between edge efficiency and advanced reasoning, nearing 26B MoE at <50% the memory footprint.
ALT Gemma 4 12B
Tris Warkentin retweeted
Introducing the newest Coral board, for efficient, on-device AI!
Check out the demos in the video:
- On-board speech translation
- Natural language controlling hardware
- Vision & sound generating music
This is perhaps the coolest demo I've seen in the last year -- 100% local intelligence that has personality and all of the live-streaming features you'd want.
What a testament to the open community -- Hugging Face building on NVIDIA hardware and models, powered by Google Gemma.
Introducing local Reachy Mini conversations: free chats forever!
So fast that we had to hardcode delays to stop it from interrupting you mid-sentence.
We built an open-source Realtime API powered by llama.cpp:
Parakeet -> Gemma 4 E4B -> Qwen3TTS
Run it anywhere you run local LMs. Video shows DGX Spark and a 36GB M3 Pro MacBook.
Blog: huggingface.co/blog/local-re…
Tris Warkentin retweeted
Here comes Google.... Maybe western "open" leadership still a possibility...
p3institute.substack.com/p/f…
Tris Warkentin retweeted
Curious what Gemma 4 can do in a local agentic setup?
This notebook uses Haystack to walk through four practical demos:
• A multimodal map/weather agent
• Dynamic tool discovery via a GitHub MCP server
• Visual Question Answering
• RAG implementations
Our new Gemini TTS model is mind-blowing -- you can use simple tags to create realistic, production-quality voice outputs. Give it a try, it's a blast!
aistudio.google.com/app/gene…
Gemma 4 is here! Performance that beats top open models at 10-20x smaller size.
One truly amazing achievement: these are the first Gemma models to achieve state-of-the-art coding and agentic capabilities as well. We are excited to see what you build with them!
Tris Warkentin retweeted
Open models are moving beyond the data center and into the physical world. 🤖
Models like @Alibaba_Qwen, @Google's Gemma, @MistralAI’s Mistral, @OpenAI’s GPT‑OSS, @physical_int’s Pi, and others now run locally on NVIDIA Jetson, powering everything from industrial machines to personal AI assistants built with OpenClaw. 🦞
See how developers are optimizing open models on Jetson to build real-time physical AI at the edge.
Read the blog 🔗 nvda.ws/4bbKH2h
