Summary

  • Gemini 1.5 Pro is a multi-modal language model with a 2-million-token context window, promising improvements across the board.
  • NotebookLM, updated with Gemini 1.5 Pro, helps students study effectively with a virtual discussion feature simulating classroom learning.
  • Users can ask questions or steer the discussion in a specific direction, making learning more interactive and realistic.

Taking the stage at the I/O developer event, Google today announced the upgraded Gemini 1.5 Pro large language model, which takes the capabilities of Google's AI even further. For one thing, Gemini 1.5 Pro, which is available today for developers, is a multi-modal language model, meaning it can operate with text, voice, and other forms of content. It also offers a 2-million-token context window, increasing the capabilities of the AI even further, and the company also promise improvements across the board to things like reasoning, translation, and more.

Using Gemini to help children study

One of the big examples Google showed on stage was from NotebookLM, a study tool that relies on AI to help children study more effectively. Updated with Gemini 1.5 Pro, NotebookLM can create a notebook guide including a study guide and quizzes to help students prepare. But the most interesting part is the new voice discussion feature.

With Gemini 1.5 Pro's multi-modal capabilities, NotebookLM can take all the information in a notebook and generate a virtual, realistic discussion talking about the topics on the notebook, simulating a more realistic experience learning in a classroom. What's more, at any time, the user can join the conversation to ask a question or steer the discussion in a specific direction. It's truly impressive stuff.

Interestingly, this all comes just after OpenAI introduced GPT-4o, which is also an impressive multi-modal LLM.

👁 GPT-4o introduced welcome banner with ChatGPT
We tried out GPT-4o, and it's so much faster than GPT-4

If you're curious how much faster GPT-4o is, the difference between it and GPT-4 is staggering.

Google Workspace with Gemini

Google also showcased some new abilities for Gemini in its Workspace suite, including Gmail and Google Meet. For example, now you can ask Gemini to summarize recent emails from a specific sender, so you if you can catch up on a project or what's going on with your kid at school, for example.

Another example includes summarizing the key points of a meeting in Google Meet, so you don't have to sit through the entire thing to get a general idea of what's being discussed. Google has been implementing AI in Workspace for some time, with features like being able to write emails for you.

This is only one subset of the many AI capabilities Google is focusing on this year. The company is all-in on artificial intelligence and competing with OpenAI, including through its new Project Astra agent.