“21/ @sundarpichai: “I’m really excited about the next phase of our partnership as we work together on agentic AI, robotics and bringing the benefits of AI to more people around the world.”” / X https://x.com/AtomSilverman/status/1902087731100209624
“The updated Google Deep Research is really good. It is much more obviously smart and agentic in its research progression, while still casting the widest net of any of the Deep Research tools. It is becoming clear that more powerful models leads to better agents for research. https://x.com/emollick/status/1900377760322785550
“20/ @NVIDIA and @Google just announced sweeping initiatives to advance agentic AI, robotics, and physical AI applications across healthcare, manufacturing, and energy sectors.” / X https://x.com/AtomSilverman/status/1902087728713625902
“Gemini Robotics model is able to generalize to tasks it has never seen before in training. https://x.com/TheHumanoidHub/status/1900262977582096576
“Meet Gemini Robotics: our latest AI models designed for a new generation of helpful robots. 🤖 Based on Gemini 2.0, they bring capabilities such as better reasoning, interactivity, dexterity and generalization into the physical world. 🧵 https://x.com/GoogleDeepMind/status/1899839624068907335
“Google dropped two AI models to advance robotics with interactivity, dexterity, and generalization: —Gemini Robotics, a VLA model, handles unfamiliar tasks like spelling words with scrabble tiles —Gemini Robotics ER brings enhanced spatial reasoning https://x.com/adcock_brett/status/1901303242459365592
“Google’s Gemini 2.0 Flash’s new image output mode has landed in Artificial Analysis Image Arena with an ELO of 1004, putting it in the middle of the leaderboard However – image output from @Google’s Gemini 2.0 Flash is still a big deal. Gemini 2.0 Flash’s integration of image https://x.com/ArtificialAnlys/status/1902038008033079326
“Gemini Canvas for coding looks great but looks like you can use it with gemini 2.0 flash for now for more control try out gemini with canvas for coding in anychat, includes support for gemini 2.0 pro and flash thinking here is a tic tac toe game with gemini 2.0 pro and canvas https://x.com/_akhaliq/status/1902039657971319110
Thoughts on Claude Code / AI workflows https://docs.google.com/presentation/d/1XRIflchTrZR2aqvxd7dROu8-76mf5N4Priwk8l9wNns/edit?slide=id.g3412c9db504_2_108#slide=id.g3412c9db504_2_108
“If you want to make images that are close to how you imagined, there are now excellent options like Gemini Multimodal (free), ideogram, flux. and Google Imogen, among others. If you want to make something interesting and strange, nothing comes close to Midjourney and its srefs. https://x.com/emollick/status/1901134192928317672
“If I guess one causality on how Google created Transformer, it’s information reversal structure in the English-Japanese translation task. (+Noam) https://x.com/shaneguML/status/1901750753548800041
“There was controversy over the way the big AI companies did image generation a year ago. Now Grok can make pictures of famous people, Gemini can remove people or watermarks from photos, etc. Is it a shift in responsibility from firms to users? A lack of negative impact? Apathy?” / X https://x.com/emollick/status/1901798161846595851
“Reskinning magic cards with Gemini (If you know the minor issue, you are a nerd) https://x.com/emollick/status/1900290565775794456
“🚀 MCP + Langflow = Next-Level AI Integration! @MisbahSy dives into Model Context Protocol (MCP)—the USB-C for AI — for how AI apps interact with tools & resources. 🛠️ See how to: 🔹 Fetch real-time web data with MCP tools 🔹 Integrate GitHub, Google Drive, Email, & more 🔹 https://x.com/langflow_ai/status/1899220193634869474
“🤯 Gemma 3’s image analysis blew me away! Tested 2 ways to extract airplane registration numbers from photos with 12B model: 1️⃣ @Gradio app w/API link (underrated feature IMO) + ZeroGPU infra on @huggingface in Google Colab. Fast & free. 2️⃣ @lmstudio server + local processing https://x.com/fdaudens/status/1900285203135987943
“With Canvas in Gemini, you can: ⌨️ Write, iterate & preview React/HTML code 📝 Draft & edit comprehensive documents 🎨 Build interactive prototypes, games & visualization …and more. Simply select ‘Canvas’ in your prompt bar and you can write and edit documents or code, with https://x.com/GeminiApp/status/1902029746508124491
“Introducing the Gemma package, a minimalistic library to use and fine-tune Gemma 🔥 Including docs on: – Fine-tuning – Sharding – LoRA – PEFT – Multimodality – Tokenization !pip install gemma https://x.com/osanseviero/status/1902456220876787763
“Photoshopping will never be the same. Gemini 2.0 Flash in a @Gradio app = 🤯 https://x.com/fdaudens/status/1901704690598842864
“I find playing with Gemini multimodal image generation to be really fun. Took a pic: “turn the bottles into a Saturn V complete with tiny ground crew. Add a neon sign to the cups saying ‘moon’ with an up arrow” “Make the rocket out of legos. Make the crew ducklings on stilts” https://x.com/emollick/status/1901370982557794658
“⚡ AutoQuant I updated AutoQuant to make the GGUF versions of Gemma 3 abliterated. It implements imatrix and can split the model into multiple files. The GGUF code is based on gguf-my-repo, maintained by @ngxson and @reach_vb It also supports GPTQ, ExLlamaV2, AWQ, and HQQ! https://x.com/maximelabonne/status/1902309252821143682
How Google Research and partners built FireSat, an early wildfire mitigation system https://blog.google/technology/ai/inside-firesat-launch-muon-space/
6 Google Health AI updates from The Check Up event 2025 https://blog.google/technology/health/the-check-up-health-ai-updates-2025/
“I tested a bunch of models on instruction following on patents yesterday. Findings: – Mistral Small 3 is better than Gemini Flash 2.0 – Mistral models are pretrained on way more patents, evident by their lower perplexity scores” / X https://x.com/casper_hansen_/status/1901540769040683214
“Using Gemini to find the source for this shows how confusing using AI for factual info is. 2.0 Flash gave a different definitive answer each time, even insisting it did a web search when it did not. Only telling it to do a reverse image search worked (but who knows if it did?) https://x.com/emollick/status/1901791796230799562
“Messing around with the new OpenAI Agents SDK but using Gemini models. I’m orchestrating a team of agents to write and check SQL statements. Can’t believe how fast, cheap and good the Gemini 2.0 models are. 🙌🏻 @OfficialLoganK @JeffDean @demishassabis” / X https://x.com/ryancarson/status/1900947290019246132
“This is wild. Google’s new Gemini model turns complex effects authoring into simple text prompts. Technical barriers gone – just describe what you want. Entire ComfyUI workflows now collapsed into simple prompts. 5 workflows you should try for free in Google AI studio. https://x.com/bilawalsidhu/status/1901078553736999340
“New skill unlocked: Gemini 2 Flash model is really awesome at removing watermarks in images! https://x.com/deedydas/status/1901042632958345369
Mind Maps – NotebookLM Help https://support.google.com/notebooklm/answer/16070070
“Using Gemini Flash Experimental to ruin art by adding ice cream. https://x.com/emollick/status/1900056829683462234
“If you have used LLM image generators, you know they are hard to control: LLM had to send a prompt to a separate image generation tool, it did not not make the image. Gemini is the first public release of a full multimodal LLM that can directly make images. Big capability gain. https://x.com/emollick/status/1899985671701311911
“Gemini can now execute code in a Canvas, a feature only Claude and ChatGPT has so far. It only does Gemini 2.0 Flash, which means it is fast (this is real-time response for: “create a reverse moonlanding game”) but also limited compared to big models like Sonnet 3.7 or GPT 4.5 https://x.com/emollick/status/1902085748146209237
“Google introduced Canvas, a collaborative space in Gemini It provides users with a code editor, code previews, and doc creation flow (similar to ChatGPT Canvas and Claude Artifacts) Google also added Audio Overviews to the Gemini app. https://x.com/rowancheung/status/1902250033493451133
New Gemini features: Canvas and Audio Overview https://blog.google/products/gemini/gemini-collaboration-features/
“Thinking for longer (e.g. o1) is only one of many axes of test-time compute. In a new @Google_AI paper, we instead focus on scaling the search axis. By just randomly sampling 200x & self-verifying, Gemini 1.5 ➡️ o1 performance. The secret: self-verification is easier at scale! https://x.com/ericzhao28/status/1901704339229732874
