Google fights OpenAI's o1 with Gemini 2.0 Flash Thinking

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 fights OpenAI's o1 with Gemini 2.0 Flash Thinking

• Google’s Gemini AI Can Summarize Your Google Drive

• AI Analysts estimate Large language models aren't so large anymore

Google
👊 Google fights OpenAI's o1 with Gemini 2.0 Flash Thinking

Google reveals Gemini 2.0 Flash Thinking, an experimental model that explicitly shows its thoughts, just like OpenAI’s 01 model does. Built on Gemini 2.0 Flash's architecture, the new model makes its problem-solving steps visible to users. The company says the algorithm can tackle problems in programming, physics, and math.

Insights for you:

• Google has introduced Gemini 2.0 Flash Thinking, an experimental AI model that offers transparency into its thought processes. The launch of “Flash Thinking” should create more competition for o1.

• Developers can now access the new Flash Thinking model through the Gemini API in Google AI Studio and Vertex AI.

• Unlike competitor reasoning models o1 from OpenAI, “Flash Thinking” enables users to access its step-by-step reasoning through a dropdown menu, offering clearer insight into how the model arrives at its conclusions.

Unlock Windsurf Editor, by Codeium.

Introducing the Windsurf Editor, the first agentic IDE. All the features you know and love from Codeium’s extensions plus new capabilities such as Cascade that act as collaborative AI agents, combining the best of copilot and agent systems. This flow state of working with AI creates a step-change in AI capability that results in truly magical moments.

Download It Free Today

Google
💡 Google’s Gemini AI Can Summarize Your Google Drive

Google is rolling out an update for Drive users that brings Gemini into its side panel with new tools. Gemini in Drive aims to help summarize a user's folders, documents, spreadsheets, and more. You can also drag and drop a folder into the Gemini sidebar, as well as right-click on a folder and choose “Ask Gemini.”

Insights for you:

• Google’s latest Gemini AI integration lets you request a summary of an entire Google Drive folder. Currently, the feature can summarize and respond to questions for text documents, PDFs, spreadsheets, and presentations. It can also respond to questions about photos.

• The response can help locate particular files in a folder or provide insights into the theme of its contents.

• To get a summary, navigate to a Google Drive folder and click the new “Summarize this folder” button. The command is currently available to paying users of Gemini AI.

AI Research
🤔 AI Analysts estimate Large language models aren't so large anymore

The conventional wisdom in AI development used to be simple: bigger models meant better performance. It's still not wrong, but it's also being challenged as the latest generation of AI language models achieves similar results with far fewer parameters.

Insights for you:

• According to new research from EpochAI, GPT-4o operates with approximately 200 billion parameters, while Claude 3.5 Sonnet uses around 400 billion parameters, compared to GPT-4, which reportedly contained 1.8 trillion parameters. These new models achieve similar or better results.

• GPT-4o, for example, generates between 100 and 150 tokens per second and costs $10 per million tokens—making it significantly faster and more cost-effective than the original GPT-4.

• The reason why this happened is that there is a higher demand for AI products as fast as can and improved "in-context reasoning" methods using synthetic data that enable smaller models to handle complex tasks effectively.