Google’s Gemma 3 AI Models beats DeepSeek

Hello, AI Enthusiasts!

Welcome to FavTutor’s AI Recap! We’ve gathered all the latest and important AI developments for the past 24 hours in one place, just for you.

In Today’s Newsletter: 😀

• Google’s Gemma 3 AI Models beats DeepSeek

• Gemini 2.0 Flash Accidentally Erases Watermarks

• Google Launches AI Research Assistant for Scientists

Google
👊 Google’s Gemma 3 AI Models beats DeepSeek

​Google has unveiled Gemma 3, a lightweight, open-source AI model supporting over 140 languages and capable of running efficiently on a single GPU. According to the official blog, Gemma 3 offers a 128,000 tokens context window, as well as supports function calling.

Insights for you:

• Gemma 3 is the latest addition to Google’s lightweight and open-source family of AI models designed to run directly on laptops and smartphones.

• This is because of its design as a single-accelerator model, which is capable of running efficiently on just one GPU or TPU.

• On the LMArena leaderboard, it has outperformed Llama-405B, DeepSeek-V3, and o3-mini by human evaluators:

Google
🧼 Oops! Gemini 2.0 Flash Accidentally Erases Watermarks

Google's Gemini 2.0 Flash AI model has sparked controversy for its ability to remove watermarks from copyrighted images. Users discovered this feature through social media posts, demonstrating how the AI seamlessly fills in gaps where watermarks once existed.

Insights for you:

• The experimental AI model of Google Gemini 2.0 Flash can remove watermarks from images, hinting at the lack of guardrails.

• Here’s how users are doing it: Go to Google AI Studio, and select “Gemini 2.0 Flash (Image Generation) Experimental” in the model picker. Then, in the prompt box, attach an image and add the prompt to remove the watermark.

• This capability raises serious ethical and legal concerns regarding intellectual property rights.

Google
🧪 Google Launches AI Research Assistant for Scientists

Google Research has created a new AI system called AI Co-Scientist that works alongside human researchers to generate and test scientific hypotheses. Built on Google's Gemini 2.0 model, the system aims to accelerate breakthrough discoveries by serving as a virtual research partner.

Insights for you:

• Google Research has developed a multi-agent AI system called AI Co-Scientist to serve as a virtual research partner.

• It iteratively generates, evaluates, and refines hypotheses to accelerate scientific discovery.

• In laboratory experiments, AI Co-Scientist has produced promising results in the development of new drugs against leukemia, the identification of new treatment targets for liver fibrosis, and the explanation of antibiotic resistance mechanisms.