Intermediate

Other Experiments

Beyond the creative FX tools, Labs.Google hosts a diverse collection of AI experiments spanning education, coding, research, and more.

Learn About

Learn About is an AI-powered tutoring experiment that helps users understand complex topics through interactive, conversational learning. Unlike a simple search engine or chatbot, Learn About is specifically designed for educational purposes.

Key Features

  • Adaptive explanations: Adjusts its explanations to your level of understanding, starting simple and adding complexity as needed.
  • Visual aids: Generates diagrams, charts, and illustrations to complement text explanations.
  • Interactive Q&A: Encourages follow-up questions and guides you through a topic step by step.
  • Multi-subject support: Covers science, history, mathematics, technology, arts, and more.
  • Source transparency: Indicates where information comes from and acknowledges uncertainty.
Learn About Example 
Topic: How does quantum computing work?

Learn About response:
Let me break this down step by step...

Classical bits are like light switches - they
are either ON (1) or OFF (0).

Quantum bits (qubits) are like a coin spinning
in the air - they can be in a state that is
both 0 AND 1 simultaneously. This is called
superposition.

Would you like me to explain superposition in
more detail, or shall we move on to how qubits
work together?
✅
Tip: Learn About works best when you engage in a conversation. Ask follow-up questions, request simpler explanations, or ask for real-world examples to deepen your understanding.

Whisk

Whisk is an image remixing experiment that lets you combine multiple images into new AI-generated visuals. Instead of writing text prompts, you provide reference images and Whisk blends their elements together.

How Whisk Works

  1. Choose a Subject Image

    Upload or select an image that defines the main subject of your creation.

  2. Choose a Scene Image

    Provide an image that defines the background, environment, or setting.

  3. Choose a Style Image

    Upload an image whose artistic style you want applied to the final result.

  4. Generate

    Whisk combines all three inputs into a new image that merges the subject, scene, and style.

Whisk is particularly useful for concept art, mood boarding, and exploring visual ideas without needing to articulate them in text.

Jules

Jules is Google's experimental AI coding agent. It is designed to help developers with coding tasks by understanding codebases, writing code, debugging, and performing code reviews.

Jules Capabilities

Feature Description
Code generation Write new code from natural language descriptions across multiple languages
Bug fixing Identify and fix bugs in existing code with explanations
Code review Review pull requests and suggest improvements for code quality
Codebase understanding Navigate and understand large codebases to answer questions about architecture
GitHub integration Works with GitHub repositories and can create pull requests
💡
Good to know: Jules integrates with GitHub and can work asynchronously on coding tasks. You can assign issues to Jules, and it will work on them independently, creating pull requests when ready for review.

Illuminate

Illuminate transforms academic papers and research documents into engaging audio discussions. It uses AI to read, understand, and then generate a podcast-style conversation about the content.

How Illuminate Works

  • Upload a paper: Provide a research paper, article, or document (typically in PDF format).
  • AI analysis: Illuminate reads and comprehends the key findings, methodology, and conclusions.
  • Audio generation: It produces a natural-sounding audio discussion between two AI voices that explain and discuss the paper's content.
  • Key takeaways: The generated discussion highlights the most important points and makes complex research accessible.

Illuminate is especially useful for researchers who want to quickly understand papers outside their specialty, students learning about new topics, or anyone who prefers audio learning.

NotebookLM

NotebookLM is one of the most successful Labs.Google experiments, having graduated into a full Google product. It is an AI-powered research and note-taking tool that lets you have conversations with your own documents.

NotebookLM Features

📄

Source Grounding

Upload documents, PDFs, websites, and videos. NotebookLM bases all its responses on your uploaded sources, reducing hallucination.

🎙

Audio Overviews

Generate podcast-style audio discussions about your sources, similar to Illuminate but integrated into a note-taking workflow.

📝

Smart Notes

Create AI-generated summaries, study guides, FAQs, and briefing documents from your source materials.

💬

Conversational Q&A

Ask questions about your uploaded sources and get answers with citations pointing back to specific passages.

✅
Tip: NotebookLM's Audio Overviews feature has become extremely popular for creating podcast-style content from documents. Upload a research paper, report, or article and generate an engaging audio discussion that you can listen to on the go.

Discovering New Experiments

Google regularly adds new experiments to Labs.Google. Here is how to stay up to date:

  • Visit Labs.Google regularly: New experiments appear on the homepage when they launch.
  • Google AI Blog: The Google AI Blog announces major new experiments and updates.
  • Google DeepMind: Follow Google DeepMind for research that may lead to new experiments.
  • Social media: Follow Google AI accounts on X (Twitter), YouTube, and LinkedIn for announcements.
  • Google I/O: Google's annual developer conference often previews upcoming Labs experiments.

Providing Feedback

Your feedback is valuable and helps shape the future of these experiments. Most experiments include feedback mechanisms:

  • Thumbs up/down: Rate individual outputs to help improve the AI models.
  • Report issues: Flag content that is inappropriate, incorrect, or problematic.
  • Feature requests: Submit suggestions for improvements or new features.
  • Surveys: Periodic surveys may appear asking about your experience with specific experiments.

Experiment Lifecycle

Understanding the lifecycle of experiments helps set expectations:

Stage Duration What Happens
Launch Day 1 Experiment becomes publicly available, often with waitlist access
Early Access Weeks to months Limited availability, rapid iteration based on early user feedback
Open Access Months Available to all eligible users, feature additions and model improvements
Maturation Months to years Stable feature set, potential integration with other Google products
Graduation Varies Becomes a standalone Google product (e.g., NotebookLM)
Archival Varies Experiment is discontinued; users are notified and given time to save their work
⚠
Important: Since experiments can be archived at any time, always download and save any AI-generated content you want to keep. Do not rely on the platform for long-term storage of your creations.

Ready to Go Deeper?

Live instructor-led courses from our partners. Affiliate disclosure.

🧠

AI & ML Courses - 30% Off

Live instructor-led AI, machine learning, data science, and cloud courses for working professionals. Use code Limited30 at checkout.

Edureka
📊

DataCamp - AI & Data Science

Hands-on Python, machine learning, and AI courses with interactive exercises and real projects.

DataCamp
🎓

edX - Top AI Courses

University-level AI courses from MIT, Harvard, Stanford. Earn certificates that employers recognize.

edX