The AI Daily Brief: "10 AI Projects to Learn Gemini 3, Nano Banana, and Opus 4.5"

Host: Nathaniel Whittemore (NLW)
Date: November 28, 2025

Episode Overview

This episode is a hands-on guide to exploring the capabilities of the latest wave of AI models—primarily Gemini 3, Nano Banana (specifically Nano Banana 2), Opus 4.5, GPT-5.1 (and its Codex Pro and Pro variants), and Grok 4.1—through ten practical projects. NLW shares concrete ideas for both technical and nontechnical listeners to experiment with these tools, better understand their strengths, and accelerate their workflows, touching on everything from image editing to auto-generating infographics, slide decks, and even entire web apps.

Tone: Practical, energetic, enthusiastic, and highly accessible.

Main Discussion Points and Insights

1. Kickoff: The Explosion of New AI Models

• NLW notes a “tear of incredible new models” in recent weeks: GPT-5.1, 5.1 Codex Pro, Gemini 3, Nano Banana 2, Opus 4.5, and Grok 4.1.
• The big takeaway: New capabilities have radically expanded what’s possible with AI tools.

2. Bonus Recommendation: Whisper Flow for Dictation

• Suggestion: "Go install Whisper Flow and try out dictation."
• NLW explains that traditional voice-to-text (on macOS/iOS, for example) is lagging, but Whisper Flow “fixes that pretty significantly.”
• Integrating speech instead of typing can speed up some workflows dramatically—NLW personally uses it at “something like 140 words a minute.”
• Notable Quote:

  "My guess is that you will very quickly find there are certain types of tasks that you will just not want to type for anymore." (04:37)

3. Project 1: Nano Banana – Infographics from Complex Content

• Highlight: Nano Banana (paired with Gemini 3) creates high-density, high-quality infographics from large content sources (like podcasts or reports).
• Example: Eric Sun’s infographic summarizing a 3.5-hour podcast episode.
• Insight: Previously, this level of information-dense visualization was “not possible at all”; now, thanks to Gemini 3’s reasoning, it’s seamless.
• Advice: Start by dropping a work report or proposal into Gemini 3 or NotebookLM and prompt for an infographic.
• Notable Quote:

  "While the first couple of weeks are just people being impressed ... I do think very quickly you're going to have a little bit of a slop sense when it comes to some of these infographics and there will be a whole bunch of human taste involved..." (10:52)

4. Project 2: Data Visualization with Gemini 3 + Nano Banana

• Use Case: Compare ideal vs actual time spent by visualizing calendar data, or turn a resume into competency Venn diagrams (idea from Zara Zhang).
• Advice: Use Gemini 3/Nano Banana to “visualize the difference between what your goals were and what you actually spent time on.”
• Notable Example:

  "She used nanobanana inside of NotebookLM to turn a resume into a slide deck ... to visualize competencies in Venn diagrams." (17:16)

5. Project 3: Advanced Image Editing with Nano Banana

• Recommendation: Edit images with Nano Banana for precise, contextual changes—such as adjusting a generated image’s style, theme, or specific elements.
• Personal Story: NLW used it to create more “Thanksgiving-y,” cartoony images for a Gratitude podcast.
• Quote:

  "Just go try to edit an image with nanobanana as compared to other image generation tools... just because it's really simple doesn't mean you should be ignoring it." (20:34)

6. Project 4: Explore New Features in NotebookLM

• Tip: Leverage NotebookLM for video overviews, explainer videos, infographics (thanks to Nano Banana), and AI-generated slide decks.
• Demo: Using NotebookLM with 22 report sources to generate a robust infographic and slide deck summarizing AI adoption patterns.
• Recommendation: Build a notebook on a topic you know, test its summarization and visualization capacities, and see how it stacks up to Gemini 3’s interface.
• Quote:

  "Although it has the look and feel of a nanobanana infographic, ... when you dig into this, the quality of information is really, really high." (26:45)

7. Project 5: Native Multimodality Reasoning Projects with Gemini 3

• Challenge: Use Gemini 3’s “integrated reasoning with native multimodality” to design a full brand system: logos, descriptions, website copy, merch, etc.
• Goal: See how far you can push single-step, multi-asset generation for a project, hobby, or business.
• Quote:

  "The thing that makes Gemini 3 so powerful is the integrated reasoning with native multimodality." (28:36)

8. Project 6: Planning and Strategy with GPT-5.1 / 5.1 Pro

• Framework: Use the new models’ advanced reasoning for “thinking out loud, thinking through strategy.”
• How-to: Provide goals and context (even through dictation); use 5.1’s auto, standard, extended, and heavy modes as needed; use Pro for synthesizing and summarizing extensive planning sessions.
• Tip: Don’t overengineer your prompt—let the model process ambiguous or partially formed objectives.
• Notable Quote:

  "I have found that5.1 is much better than previous ChatGPT models at just making a decision and making an argument for it." (35:20)

9. Projects 7-9: Vibe Coding with No-Code/Low-Code Platforms

• Emergence: Platforms like Replit, Lovable, and Google AI Studio are supercharged by new models, unlocking full-stack, end-to-end website and app development for non-developers.
• Challenge: Turn your 2026 strategic plan into a personal accountability web app—complete with timeline, goal check-ins, file uploads, and automated reminders.
• Advanced: Integrate AI-generated infographics, chatbots, conversational agents, and animated graphics by tying in Gemini API and Nano Banana features.
• Visual Leap: Try Replit’s new design mode, powered by recent models, to break away from dated, generic templates.

10. Project 10: Deep Coding and Complex Tasks with Opus 4.5

• Note: For those more technically inclined, Opus 4.5 is “pretty much the best coding model” so far, excelling at not getting “lost in the sauce of deep coding tasks.”
• App Ideas: Build a content repurposing hub: input a transcript, generate a LinkedIn article, social posts, newsletter, tweet thread, and extract key quotes automatically.
• Bonus: Claude for Excel—excel in complex spreadsheet-based tasks.

Notable Quotes & Memorable Moments

• On the difference between 'just trying things' and expertise:

  "Don’t be afraid to just try things out to start. You can always go back and edit later." (09:35)

• On evolving standards for AI-generated infographics:

  "There will be a whole bunch of human taste involved in nudging the model in directions that makes the visual presentation not just the sort of default nanobanana setting." (11:15)

• On Vibe Coding platforms' new accessibility:

  "For the non technical vibe coders that couldn't speak in code until 2025 ... you can now with any of the majors ... build a website that is actually published that you and others can actually interact with." (41:30)

• On Opus 4.5’s leap for coders:

  "All the people who are using it seem to be emphatic that it is pretty much the best coding model they've ever used." (46:40)

Key Timestamps

• [04:37] – Whisper Flow recommendation for dictation
• [10:52] – Nano Banana, Gemini 3, and the future of infographics
• [17:16] – Data visualization applications (resume to Venn diagram)
• [20:34] – Advanced image editing with Nano Banana
• [26:45] – NotebookLM’s new video/explainer/infographic features
• [28:36] – Defining power of Gemini 3’s reasoning and multimodality
• [35:20] – GPT-5.1’s superiority at decision-making in strategy generation
• [41:30] – Vibe Coding is now truly end-to-end, even for nondevs
• [46:40] – Opus 4.5 sets a new bar for coding LLMs

Final Takeaway

Rather than just observing the headline capabilities, NLW urges listeners to “go create, experiment, and publish” with these new tools:

"There have been so many models that even 10 AI projects probably can’t encapsulate all the new things you can do, but hopefully that gives you some ideas of where you can dive in...” (48:02)

For the next few days, try out these projects, share what you build, and see just how far the new wave of AI can take you—both as a creator and a strategic thinker.

End of actionable summary.